=== 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: 1df1199c145652d4a172ffedffd9b9e59ce98d27 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 04.10.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/01 job id: 49980676 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/02 job id: 49980677 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/03 job id: 49980678 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/04 job id: 49980680 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/05 job id: 49980681 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/06 job id: 49980682 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/07 job id: 49980683 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/08 job id: 49980684 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/09 job id: 49980685 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/10 job id: 49980686 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/11 job id: 49980687 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/12 job id: 49980688 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/13 job id: 49980689 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/14 job id: 49980690 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/15 job id: 49980691 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/16 job id: 49980692 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/17 job id: 49980693 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/18 job id: 49980694 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/19 job id: 49980695 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/20 job id: 49980696 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/21 job id: 49980697 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/22 job id: 49980698 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/23 job id: 49980700 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/24 job id: 49980701 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/25 job id: 49980702 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/26 job id: 49980703 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/27 job id: 49980704 --- 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/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 0.018 0.033 133.122 133.123 farming_run 1 2.0 132.559 132.561 133.093 133.096 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.492361E+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 0.005 0.025 114.761 114.761 qs_energies 1 2.0 0.000 0.000 114.568 114.571 mp2_main 1 3.0 0.000 0.000 112.509 112.512 mp2_gpw_main 1 4.0 0.033 0.039 111.520 111.523 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 91.949 92.300 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.727 55.079 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.165 41.140 46.211 get_2c_integrals 1 6.0 0.008 0.009 36.441 37.057 integrate_v_rspace 273 8.0 0.441 0.458 24.694 29.570 pw_transfer 6555 10.6 0.377 0.389 26.836 27.326 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 25.479 26.034 grid_integrate_task_list 273 9.0 20.585 25.939 20.585 25.939 fft_wrap_pw1pw2_100 2178 12.4 0.028 0.030 23.029 23.566 rpa_ri_compute_en 1 5.0 0.020 0.023 19.460 19.682 compute_2c_integrals 1 7.0 0.002 0.002 18.876 18.883 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.541 18.610 mp2_eri_2c_integrate_gpw 1 9.0 2.389 2.425 18.539 18.609 cp_fm_cholesky_decompose 12 8.2 17.479 18.135 17.479 18.135 cholesky_decomp 1 7.0 0.000 0.000 16.404 17.067 fft3d_s 5443 13.4 16.129 16.581 16.150 16.602 ao_to_mo_and_store_B_mult_1 272 7.0 10.765 15.313 10.765 15.313 calculate_wavefunction 272 8.0 5.414 5.573 12.342 13.021 rpa_num_int 1 6.0 0.000 0.004 10.584 10.584 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.575 10.580 calc_mat_Q 8 8.0 0.000 0.000 9.455 9.534 contract_S_to_Q 8 9.0 0.000 0.000 8.876 8.953 calc_potential_gpw 544 9.5 0.004 0.005 8.242 8.671 parallel_gemm_fm 14 9.1 0.000 0.000 8.459 8.558 parallel_gemm_fm_cosma 14 10.1 8.459 8.558 8.459 8.558 create_integ_mat 1 6.0 0.025 0.033 8.431 8.431 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.044 8.270 potential_pw2rs 545 10.0 0.106 0.108 7.470 8.216 array2fm 1 7.0 0.000 0.000 7.157 7.848 collocate_single_gaussian 272 10.0 0.039 0.041 7.302 7.553 pw_scatter_s 2720 13.7 4.351 4.495 4.351 4.495 array2fm_buffer_send 1 8.0 3.017 3.913 3.017 3.913 pw_gather_s 2722 13.2 3.470 3.828 3.470 3.828 array2fm_fill 1 8.0 0.859 0.880 1.739 2.484 pw_poisson_solve 545 10.5 1.117 1.167 2.180 2.333 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.522890, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2798.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.035 421.495 421.497 farming_run 1 2.0 420.637 420.644 421.455 421.458 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.248829E+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 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.032 217.402 217.403 qs_energies 1 2.0 0.002 0.016 217.187 217.194 scf_env_do_scf 1 3.0 0.000 0.000 115.083 115.084 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.199 114.205 rebuild_ks_matrix 4 6.0 0.000 0.000 114.198 114.204 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 114.198 114.204 hfx_ks_matrix 4 8.0 0.001 0.001 113.855 113.860 integrate_four_center 4 9.0 0.154 0.473 113.854 113.859 integrate_four_center_main 4 10.0 0.138 0.614 101.788 105.173 integrate_four_center_bin 265 11.0 101.649 105.159 101.649 105.159 mp2_main 1 3.0 0.005 0.043 101.817 101.826 mp2_gpw_main 1 4.0 0.053 0.120 100.957 100.971 init_scf_loop 1 4.0 0.000 0.000 97.098 97.098 mp2_ri_gpw_compute_in 1 5.0 0.066 0.081 73.984 74.963 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 53.757 54.750 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 41.429 46.234 integrate_v_rspace 95 8.0 0.398 0.565 27.904 32.563 pw_transfer 2240 10.6 0.143 0.170 29.498 29.847 fft_wrap_pw1pw2 1868 11.4 0.017 0.021 28.509 28.877 mp2_ri_gpw_compute_en 1 5.0 0.055 0.064 26.751 28.305 grid_integrate_task_list 95 9.0 23.325 28.234 23.325 28.234 ao_to_mo_and_store_B_mult_1 91 7.0 10.656 27.006 10.656 27.006 fft_wrap_pw1pw2_100 730 12.4 0.012 0.013 26.261 26.661 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.836 1.901 25.101 25.111 get_2c_integrals 1 6.0 0.012 0.097 20.051 20.164 compute_2c_integrals 1 7.0 0.003 0.005 19.009 19.015 fft3d_s 1823 13.4 18.478 18.925 18.492 18.939 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.645 18.892 mp2_eri_2c_integrate_gpw 1 9.0 1.736 1.849 18.645 18.890 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.984 17.984 calculate_wavefunction 91 8.0 2.013 2.042 9.616 9.813 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.581 8.809 9.520 potential_pw2rs 186 10.0 0.033 0.034 8.415 8.944 local_gemm 172 8.0 8.251 8.938 8.251 8.938 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.000 0.001 8.166 8.572 mp2_ri_gpw_compute_en_comm 22 7.0 0.501 0.526 8.085 8.458 calc_potential_gpw 182 9.5 0.002 0.002 7.916 8.154 collocate_single_gaussian 91 10.0 0.017 0.026 7.751 8.006 mp_sync 37 10.5 4.116 7.229 4.116 7.229 integrate_four_center_load 4 10.0 0.000 0.000 6.766 6.770 hfx_load_balance 1 11.0 0.000 0.000 6.766 6.770 mp_sendrecv_dm3 2068 8.0 6.111 6.503 6.111 6.503 mp2_ri_gpw_compute_en_ener 172 7.0 6.327 6.401 6.327 6.401 pw_gather_s 912 13.2 4.475 4.964 4.475 4.964 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.945774, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1508.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 452.866048E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.025 53.655 53.656 qs_mol_dyn_low 1 2.0 0.003 0.003 53.450 53.458 qs_forces 11 3.9 0.002 0.003 53.390 53.391 qs_energies 11 4.9 0.001 0.002 51.895 51.920 scf_env_do_scf 11 5.9 0.001 0.004 46.051 46.051 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.971 43.971 qs_scf_new_mos 108 7.5 0.000 0.001 34.043 34.360 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.042 34.360 dbcsr_multiply_generic 2286 12.5 0.094 0.099 33.582 33.995 ot_scf_mini 108 9.5 0.002 0.002 32.393 32.558 multiply_cannon 2286 13.5 0.189 0.200 26.157 27.687 multiply_cannon_loop 2286 14.5 1.858 1.955 25.511 27.056 velocity_verlet 10 3.0 0.001 0.002 26.171 26.172 ot_mini 108 10.5 0.001 0.001 19.541 19.761 qs_ot_get_derivative 108 11.5 0.001 0.001 16.460 16.657 mp_waitall_1 245248 16.5 8.395 14.484 8.395 14.484 multiply_cannon_metrocomm3 54864 15.5 0.073 0.077 6.023 12.536 multiply_cannon_multrec 54864 15.5 3.625 5.690 7.624 11.001 qs_ot_get_p 119 10.4 0.001 0.001 8.239 8.520 rebuild_ks_matrix 119 8.3 0.000 0.000 7.784 7.911 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.784 7.911 mp_sum_l 7287 12.8 5.512 7.115 5.512 7.115 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.849 6.960 multiply_cannon_sync_h2d 54864 15.5 5.113 6.290 5.113 6.290 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.607 6.057 dbcsr_mm_accdrv_process 76910 16.1 1.871 2.920 3.913 5.586 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.386 5.493 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.246 5.287 init_scf_run 11 5.9 0.000 0.001 4.584 4.585 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.584 4.584 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.445 4.445 sum_up_and_integrate 119 10.3 0.001 0.002 4.389 4.394 integrate_v_rspace 119 11.3 0.002 0.002 4.378 4.384 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.259 4.259 cp_fm_redistribute_end 50 14.0 2.172 4.234 2.178 4.237 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.082 4.185 calculate_rho_elec 119 8.7 0.011 0.016 4.082 4.185 cp_fm_diag_elpa_base 50 14.0 2.053 4.128 2.057 4.136 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.925 3.110 apply_single 119 13.6 0.000 0.000 2.925 3.110 calculate_dm_sparse 119 9.5 0.000 0.000 2.871 3.104 multiply_cannon_metrocomm1 54864 15.5 0.056 0.062 1.744 2.857 ot_diis_step 108 11.5 0.006 0.006 2.754 2.755 acc_transpose_blocks 54864 15.5 0.226 0.247 2.236 2.742 jit_kernel_multiply 13 15.8 1.975 2.577 1.975 2.577 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.428 2.496 calculate_first_density_matrix 1 7.0 0.000 0.000 2.345 2.357 density_rs2pw 119 9.7 0.004 0.004 2.107 2.195 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.161 2.163 wfi_extrapolate 11 7.9 0.001 0.001 2.118 2.118 grid_integrate_task_list 119 12.3 2.001 2.093 2.001 2.093 mp_sum_d 4135 12.0 1.465 2.072 1.465 2.072 init_scf_loop 11 6.9 0.000 0.000 2.062 2.063 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.873 1.940 potential_pw2rs 119 12.3 0.004 0.004 1.780 1.790 pw_transfer 1439 11.6 0.052 0.056 1.624 1.690 make_m2s 4572 13.5 0.053 0.056 1.596 1.649 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.549 1.619 make_images 4572 14.5 0.133 0.139 1.514 1.567 mp_alltoall_d11v 2130 13.8 1.287 1.485 1.287 1.485 acc_transpose_blocks_sync 164592 16.5 1.239 1.470 1.239 1.470 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.389 1.427 grid_collocate_task_list 119 9.7 1.348 1.425 1.348 1.425 transfer_rs2pw 487 10.6 0.005 0.006 1.329 1.412 mp_waitany 12084 13.8 1.247 1.404 1.247 1.404 fft3d_ps 1201 14.6 0.367 0.474 1.261 1.326 transfer_pw2rs 487 13.2 0.006 0.006 1.305 1.312 fft_wrap_pw1pw2_140 487 13.2 0.107 0.113 1.201 1.271 dbcsr_dot_sd 1205 11.9 0.049 0.058 0.815 1.193 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.656000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.636364, yerr=0.642824 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 489.267200E+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 987768. 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 0.007 0.023 38.576 38.577 qs_mol_dyn_low 1 2.0 0.003 0.003 38.405 38.413 qs_forces 11 3.9 0.002 0.003 38.344 38.345 qs_energies 11 4.9 0.001 0.001 36.687 36.692 scf_env_do_scf 11 5.9 0.005 0.013 31.434 31.435 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.014 29.015 dbcsr_multiply_generic 2286 12.5 0.101 0.106 22.092 22.428 qs_scf_new_mos 108 7.5 0.001 0.001 20.497 20.727 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.497 20.727 ot_scf_mini 108 9.5 0.002 0.003 19.579 19.744 multiply_cannon 2286 13.5 0.208 0.217 17.005 18.410 velocity_verlet 10 3.0 0.001 0.001 18.139 18.140 multiply_cannon_loop 2286 14.5 1.202 1.267 15.860 17.411 ot_mini 108 10.5 0.001 0.001 12.093 12.325 mp_waitall_1 200699 16.5 5.475 10.782 5.475 10.782 qs_ot_get_derivative 108 11.5 0.001 0.001 9.738 9.902 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.010 9.563 multiply_cannon_multrec 27432 15.5 1.812 4.232 6.290 9.044 rebuild_ks_matrix 119 8.3 0.000 0.000 6.702 6.853 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.702 6.853 dbcsr_mm_accdrv_process 47894 16.0 3.660 6.020 4.398 6.529 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.895 6.031 qs_ot_get_p 119 10.4 0.001 0.003 4.587 4.804 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.655 4.564 init_scf_run 11 5.9 0.000 0.001 4.071 4.071 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.070 4.071 mp_sum_l 7287 12.8 2.070 4.068 2.070 4.068 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.060 4.040 apply_single 119 13.6 0.000 0.000 3.060 4.040 sum_up_and_integrate 119 10.3 0.001 0.001 3.642 3.648 integrate_v_rspace 119 11.3 0.002 0.003 3.628 3.634 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.559 3.590 calculate_rho_elec 119 8.7 0.021 0.024 3.559 3.590 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.950 2.968 make_m2s 4572 13.5 0.052 0.053 2.562 2.896 make_images 4572 14.5 0.205 0.243 2.473 2.809 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.552 2.552 calculate_first_density_matrix 1 7.0 0.000 0.000 2.458 2.459 init_scf_loop 11 6.9 0.000 0.000 2.396 2.401 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.295 2.380 ot_diis_step 108 11.5 0.011 0.011 2.308 2.308 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.245 2.245 multiply_cannon_sync_h2d 27432 15.5 1.681 2.234 1.681 2.234 cp_fm_redistribute_end 50 14.0 1.138 2.222 1.141 2.224 calculate_dm_sparse 119 9.5 0.000 0.001 2.142 2.220 cp_fm_diag_elpa_base 50 14.0 1.051 2.137 1.079 2.175 acc_transpose_blocks 27432 15.5 0.116 0.122 1.647 1.994 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.953 1.955 grid_integrate_task_list 119 12.3 1.842 1.936 1.842 1.936 density_rs2pw 119 9.7 0.004 0.004 1.857 1.922 jit_kernel_multiply 9 16.4 0.679 1.855 0.679 1.855 pw_transfer 1439 11.6 0.064 0.068 1.805 1.838 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.753 1.791 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.714 1.750 make_images_data 4572 15.5 0.047 0.053 1.220 1.573 prepare_preconditioner 11 7.9 0.000 0.000 1.502 1.528 make_preconditioner 11 8.9 0.000 0.000 1.502 1.528 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.403 1.460 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.083 1.438 wfi_extrapolate 11 7.9 0.001 0.001 1.432 1.432 fft_wrap_pw1pw2_140 487 13.2 0.128 0.132 1.373 1.409 potential_pw2rs 119 12.3 0.006 0.006 1.394 1.399 fft3d_ps 1201 14.6 0.521 0.576 1.344 1.373 grid_collocate_task_list 119 9.7 1.291 1.348 1.291 1.348 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.331 1.338 mp_alltoall_d11v 2130 13.8 1.182 1.308 1.182 1.308 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.193 1.240 transfer_rs2pw 487 10.6 0.005 0.005 1.045 1.123 mp_sum_d 4135 12.0 0.529 1.013 0.529 1.013 mp_allgather_i34 2286 14.5 0.569 0.966 0.569 0.966 acc_transpose_blocks_sync 82296 16.5 0.829 0.953 0.829 0.953 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.926 0.927 acc_transpose_blocks_kernels 27432 16.5 0.189 0.278 0.677 0.894 transfer_pw2rs 487 13.2 0.004 0.005 0.873 0.878 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.863 0.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.577000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.636364, yerr=1.553455 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 520.396800E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.048 31.713 31.714 qs_mol_dyn_low 1 2.0 0.003 0.004 31.423 31.433 qs_forces 11 3.9 0.003 0.008 31.358 31.359 qs_energies 11 4.9 0.002 0.006 29.812 29.814 scf_env_do_scf 11 5.9 0.001 0.003 25.190 25.190 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 22.772 22.773 dbcsr_multiply_generic 2286 12.5 0.096 0.099 16.757 16.858 qs_scf_new_mos 108 7.5 0.001 0.001 15.224 15.238 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.224 15.237 velocity_verlet 10 3.0 0.002 0.002 14.911 14.913 ot_scf_mini 108 9.5 0.002 0.003 14.463 14.474 multiply_cannon 2286 13.5 0.194 0.198 13.617 14.309 multiply_cannon_loop 2286 14.5 0.871 0.911 12.856 13.572 ot_mini 108 10.5 0.001 0.001 8.973 8.987 qs_ot_get_derivative 108 11.5 0.001 0.001 7.493 7.505 multiply_cannon_multrec 18288 15.5 1.873 2.786 6.972 7.224 rebuild_ks_matrix 119 8.3 0.000 0.000 5.915 5.936 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.018 5.914 5.936 dbcsr_mm_accdrv_process 38222 16.0 4.947 5.791 5.006 5.857 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.207 5.226 sum_up_and_integrate 119 10.3 0.001 0.002 3.495 3.501 integrate_v_rspace 119 11.3 0.003 0.003 3.482 3.490 init_scf_run 11 5.9 0.000 0.001 3.431 3.431 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.431 3.431 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.829 3.410 mp_waitall_1 158411 16.6 2.513 3.321 2.513 3.321 qs_ot_get_p 119 10.4 0.001 0.002 3.278 3.303 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.220 3.228 calculate_rho_elec 119 8.7 0.031 0.031 3.219 3.228 init_scf_loop 11 6.9 0.004 0.025 2.400 2.401 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.977 2.341 apply_single 119 13.6 0.000 0.000 1.977 2.341 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.472 2.268 calculate_first_density_matrix 1 7.0 0.001 0.004 2.201 2.203 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.180 2.189 grid_integrate_task_list 119 12.3 1.809 1.921 1.809 1.921 cp_dbcsr_syevd 50 12.0 0.003 0.006 1.885 1.885 make_m2s 4572 13.5 0.044 0.046 1.739 1.882 density_rs2pw 119 9.7 0.004 0.006 1.786 1.863 calculate_dm_sparse 119 9.5 0.000 0.001 1.847 1.861 pw_transfer 1439 11.6 0.065 0.068 1.836 1.845 make_images 4572 14.5 0.191 0.205 1.653 1.795 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.743 1.754 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.696 1.703 acc_transpose_blocks 18288 15.5 0.081 0.083 1.636 1.698 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.657 1.659 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.648 1.658 cp_fm_diag_elpa_base 50 14.0 1.624 1.638 1.645 1.655 prepare_preconditioner 11 7.9 0.000 0.000 1.605 1.609 make_preconditioner 11 8.9 0.000 0.002 1.605 1.609 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.468 1.551 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.499 1.505 ot_diis_step 108 11.5 0.011 0.011 1.463 1.463 mp_sum_l 7287 12.8 1.020 1.439 1.020 1.439 fft_wrap_pw1pw2_140 487 13.2 0.180 0.186 1.420 1.430 grid_collocate_task_list 119 9.7 1.234 1.340 1.234 1.340 fft3d_ps 1201 14.6 0.546 0.567 1.296 1.308 potential_pw2rs 119 12.3 0.007 0.008 1.297 1.301 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.181 1.184 multiply_cannon_sync_h2d 18288 15.5 1.023 1.174 1.023 1.174 wfi_extrapolate 11 7.9 0.001 0.001 1.147 1.147 transfer_rs2pw 487 10.6 0.005 0.005 0.950 1.062 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.962 0.963 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.937 0.954 make_images_data 4572 15.5 0.047 0.051 0.745 0.910 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.648 0.832 acc_transpose_blocks_sync 54864 16.5 0.749 0.811 0.749 0.811 acc_transpose_blocks_kernels 18288 16.5 0.220 0.227 0.787 0.795 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.778 0.781 transfer_pw2rs 487 13.2 0.004 0.004 0.775 0.777 mp_alltoall_d11v 2130 13.8 0.636 0.759 0.636 0.759 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.678 0.745 cp_fm_cholesky_invert 11 10.9 0.674 0.678 0.674 0.678 mp_alltoall_z22v 1201 16.6 0.602 0.675 0.602 0.675 mp_waitany 9880 13.7 0.527 0.650 0.527 0.650 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.714000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.000000, yerr=1.044466 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 557.449216E+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 0.015 0.032 36.014 36.015 qs_mol_dyn_low 1 2.0 0.003 0.003 35.832 35.841 qs_forces 11 3.9 0.002 0.002 35.773 35.774 qs_energies 11 4.9 0.001 0.002 34.071 34.077 scf_env_do_scf 11 5.9 0.001 0.002 29.086 29.088 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.755 25.755 dbcsr_multiply_generic 2286 12.5 0.103 0.106 19.794 19.916 velocity_verlet 10 3.0 0.002 0.002 18.317 18.319 qs_scf_new_mos 108 7.5 0.001 0.001 17.727 17.783 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.727 17.783 ot_scf_mini 108 9.5 0.002 0.003 16.687 16.742 multiply_cannon 2286 13.5 0.219 0.228 16.062 16.611 multiply_cannon_loop 2286 14.5 1.566 1.631 15.120 15.533 ot_mini 108 10.5 0.001 0.001 10.279 10.344 multiply_cannon_multrec 27432 15.5 2.447 3.136 9.006 9.311 qs_ot_get_derivative 108 11.5 0.001 0.001 8.409 8.463 dbcsr_mm_accdrv_process 47916 15.9 5.802 7.401 6.456 7.814 rebuild_ks_matrix 119 8.3 0.000 0.000 6.295 6.349 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.295 6.349 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.571 5.620 init_scf_run 11 5.9 0.000 0.001 3.663 3.663 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.663 3.663 qs_ot_get_p 119 10.4 0.001 0.001 3.551 3.625 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.022 3.525 sum_up_and_integrate 119 10.3 0.001 0.001 3.457 3.463 integrate_v_rspace 119 11.3 0.003 0.003 3.446 3.452 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.334 3.372 calculate_rho_elec 119 8.7 0.040 0.046 3.333 3.372 init_scf_loop 11 6.9 0.000 0.000 3.313 3.314 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.205 2.667 apply_single 119 13.6 0.000 0.000 2.205 2.667 acc_transpose_blocks 27432 15.5 0.122 0.127 2.445 2.571 prepare_preconditioner 11 7.9 0.000 0.000 2.490 2.500 make_preconditioner 11 8.9 0.000 0.000 2.490 2.500 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.096 2.424 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.327 2.355 make_m2s 4572 13.5 0.054 0.056 2.240 2.351 calculate_first_density_matrix 1 7.0 0.000 0.000 2.294 2.296 make_images 4572 14.5 0.273 0.336 2.132 2.243 mp_waitall_1 137007 16.6 1.621 2.211 1.621 2.211 calculate_dm_sparse 119 9.5 0.000 0.000 2.135 2.189 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.154 2.166 pw_transfer 1439 11.6 0.064 0.068 1.967 2.001 grid_integrate_task_list 119 12.3 1.839 1.920 1.839 1.920 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.876 1.912 density_rs2pw 119 9.7 0.003 0.004 1.789 1.909 ot_diis_step 108 11.5 0.012 0.012 1.826 1.826 jit_kernel_multiply 10 15.5 0.587 1.819 0.587 1.819 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.813 1.814 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.750 1.750 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.638 1.654 mp_sum_l 7287 12.8 1.065 1.653 1.065 1.653 fft_wrap_pw1pw2_140 487 13.2 0.228 0.234 1.562 1.599 acc_transpose_blocks_sync 82296 16.5 1.471 1.594 1.471 1.594 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.517 1.525 cp_fm_diag_elpa_base 50 14.0 1.483 1.499 1.515 1.523 fft3d_ps 1201 14.6 0.590 0.652 1.363 1.387 grid_collocate_task_list 119 9.7 1.249 1.343 1.249 1.343 wfi_extrapolate 11 7.9 0.001 0.001 1.316 1.316 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.293 1.312 potential_pw2rs 119 12.3 0.008 0.009 1.264 1.266 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.222 1.233 multiply_cannon_metrocomm3 27432 15.5 0.040 0.040 0.692 1.228 cp_fm_upper_to_full 72 14.2 0.818 1.165 0.818 1.165 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.065 1.066 dbcsr_complete_redistribute 329 12.2 0.124 0.158 0.756 1.025 transfer_rs2pw 487 10.6 0.004 0.005 0.875 1.005 make_images_data 4572 15.5 0.048 0.053 0.830 0.947 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.706 0.881 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.799 0.874 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.564 0.833 acc_transpose_blocks_kernels 27432 16.5 0.272 0.278 0.823 0.831 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.812 0.818 mp_alltoall_d11v 2130 13.8 0.709 0.772 0.709 0.772 mp_alltoall_i22 627 13.8 0.428 0.724 0.428 0.724 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.015000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.272727, yerr=2.956992 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 615.325696E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.029 28.110 28.111 qs_mol_dyn_low 1 2.0 0.003 0.003 27.881 27.915 qs_forces 11 3.9 0.002 0.003 27.818 27.819 qs_energies 11 4.9 0.001 0.002 26.124 26.126 scf_env_do_scf 11 5.9 0.001 0.001 21.498 21.498 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.094 19.094 velocity_verlet 10 3.0 0.002 0.002 14.499 14.502 dbcsr_multiply_generic 2286 12.5 0.096 0.099 12.777 12.845 qs_scf_new_mos 108 7.5 0.001 0.001 11.473 11.497 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.472 11.496 ot_scf_mini 108 9.5 0.002 0.002 10.765 10.785 multiply_cannon 2286 13.5 0.223 0.232 10.275 10.782 multiply_cannon_loop 2286 14.5 0.647 0.679 9.371 9.508 ot_mini 108 10.5 0.001 0.001 6.134 6.160 multiply_cannon_multrec 9144 15.5 1.647 1.858 5.909 6.097 rebuild_ks_matrix 119 8.3 0.000 0.000 5.743 5.768 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 5.742 5.767 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.109 5.132 qs_ot_get_derivative 108 11.5 0.001 0.001 4.856 4.875 dbcsr_mm_accdrv_process 12550 15.8 3.503 4.179 4.151 4.233 sum_up_and_integrate 119 10.3 0.001 0.001 3.394 3.402 integrate_v_rspace 119 11.3 0.003 0.003 3.384 3.392 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.339 3.344 calculate_rho_elec 119 8.7 0.060 0.061 3.339 3.344 init_scf_run 11 5.9 0.000 0.001 3.198 3.199 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.198 3.198 qs_ot_get_p 119 10.4 0.001 0.002 2.764 2.795 init_scf_loop 11 6.9 0.000 0.000 2.386 2.387 pw_transfer 1439 11.6 0.065 0.067 2.072 2.081 calculate_first_density_matrix 1 7.0 0.000 0.000 2.079 2.081 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.979 1.988 grid_integrate_task_list 119 12.3 1.867 1.927 1.867 1.927 make_m2s 4572 13.5 0.033 0.035 1.749 1.904 make_images 4572 14.5 0.269 0.302 1.660 1.812 density_rs2pw 119 9.7 0.003 0.003 1.715 1.809 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.803 1.805 calculate_dm_sparse 119 9.5 0.000 0.000 1.740 1.756 mp_waitall_1 115863 16.7 1.231 1.737 1.231 1.737 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.682 1.702 prepare_preconditioner 11 7.9 0.000 0.000 1.669 1.675 make_preconditioner 11 8.9 0.000 0.000 1.669 1.675 fft_wrap_pw1pw2_140 487 13.2 0.325 0.333 1.650 1.661 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.561 1.586 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.560 1.561 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.512 1.513 acc_transpose_blocks 9144 15.5 0.043 0.045 1.427 1.477 grid_collocate_task_list 119 9.7 1.296 1.387 1.296 1.387 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.354 1.363 fft3d_ps 1201 14.6 0.647 0.660 1.330 1.341 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.328 1.340 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.272 1.280 cp_fm_diag_elpa_base 50 14.0 1.246 1.262 1.270 1.279 ot_diis_step 108 11.5 0.012 0.013 1.265 1.265 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.211 1.238 apply_single 119 13.6 0.000 0.000 1.211 1.238 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.220 1.221 potential_pw2rs 119 12.3 0.010 0.011 1.208 1.211 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.127 1.133 wfi_extrapolate 11 7.9 0.001 0.001 1.067 1.067 jit_kernel_multiply 6 15.7 0.608 1.020 0.608 1.020 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.757 0.967 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.873 0.925 make_images_data 4572 15.5 0.042 0.046 0.749 0.905 mp_alltoall_d11v 2130 13.8 0.790 0.893 0.790 0.893 cp_fm_cholesky_invert 11 10.9 0.824 0.827 0.824 0.827 transfer_rs2pw 487 10.6 0.004 0.005 0.742 0.825 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.810 0.816 acc_transpose_blocks_sync 27432 16.5 0.728 0.777 0.728 0.777 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.753 0.755 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.666 0.719 acc_transpose_blocks_kernels 9144 16.5 0.119 0.121 0.639 0.648 mp_allgather_i34 2286 14.5 0.220 0.626 0.220 0.626 transfer_pw2rs 487 13.2 0.003 0.004 0.623 0.624 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.313 0.606 mp_alltoall_z22v 1201 16.6 0.557 0.591 0.557 0.591 qs_create_task_list 11 7.9 0.000 0.001 0.535 0.563 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.111000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=584.181818, yerr=4.018552 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 782.352384E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.031 42.398 42.398 qs_mol_dyn_low 1 2.0 0.003 0.003 42.192 42.199 qs_forces 11 3.9 0.002 0.002 42.124 42.125 qs_energies 11 4.9 0.002 0.003 40.133 40.136 scf_env_do_scf 11 5.9 0.001 0.001 34.371 34.371 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.677 26.680 velocity_verlet 10 3.0 0.002 0.008 23.902 23.908 dbcsr_multiply_generic 2286 12.5 0.104 0.106 18.840 19.041 qs_scf_new_mos 108 7.5 0.001 0.001 17.034 17.135 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.034 17.134 ot_scf_mini 108 9.5 0.002 0.002 15.878 15.984 multiply_cannon 2286 13.5 0.301 0.309 15.005 15.864 multiply_cannon_loop 2286 14.5 0.859 0.883 13.761 14.664 ot_mini 108 10.5 0.001 0.001 9.669 9.790 multiply_cannon_multrec 9144 15.5 3.378 4.772 8.791 8.869 qs_ot_get_derivative 108 11.5 0.001 0.001 7.623 7.729 init_scf_loop 11 6.9 0.000 0.000 7.668 7.675 rebuild_ks_matrix 119 8.3 0.000 0.000 7.236 7.379 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.235 7.378 prepare_preconditioner 11 7.9 0.000 0.000 6.704 6.717 make_preconditioner 11 8.9 0.000 0.000 6.704 6.717 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.524 6.655 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.298 6.599 dbcsr_mm_accdrv_process 12550 15.8 4.680 6.487 5.275 6.555 cp_fm_upper_to_full 72 14.2 3.169 4.523 3.169 4.523 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.172 4.175 calculate_rho_elec 119 8.7 0.117 0.120 4.171 4.175 sum_up_and_integrate 119 10.3 0.001 0.001 3.822 3.829 integrate_v_rspace 119 11.3 0.003 0.004 3.812 3.818 init_scf_run 11 5.9 0.000 0.001 3.665 3.665 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.665 3.665 qs_ot_get_p 119 10.4 0.002 0.002 3.294 3.435 mp_waitall_1 94719 16.7 2.211 3.214 2.211 3.214 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.786 3.203 pw_transfer 1439 11.6 0.068 0.069 3.001 3.006 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.903 2.907 dbcsr_complete_redistribute 329 12.2 0.281 0.287 1.925 2.735 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.209 2.511 apply_single 119 13.6 0.000 0.000 2.209 2.511 fft_wrap_pw1pw2_140 487 13.2 0.655 0.657 2.488 2.494 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.633 2.445 make_m2s 4572 13.5 0.037 0.037 2.283 2.432 make_images 4572 14.5 0.354 0.388 2.161 2.311 calculate_dm_sparse 119 9.5 0.000 0.000 2.247 2.264 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.337 2.240 density_rs2pw 119 9.7 0.003 0.003 2.216 2.234 mp_alltoall_i22 627 13.8 1.442 2.229 1.442 2.229 calculate_first_density_matrix 1 7.0 0.000 0.000 2.211 2.213 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.401 2.206 grid_integrate_task_list 119 12.3 2.075 2.116 2.075 2.116 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.032 2.085 ot_diis_step 108 11.5 0.014 0.014 2.017 2.017 qs_ot_p2m_diag 50 11.0 0.044 0.044 1.929 1.931 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.913 1.914 acc_transpose_blocks 9144 15.5 0.045 0.046 1.799 1.843 fft3d_ps 1201 14.6 0.867 0.885 1.818 1.823 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.815 1.816 mp_sum_l 7287 12.8 1.001 1.768 1.001 1.768 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.605 1.627 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.575 1.575 grid_collocate_task_list 119 9.7 1.518 1.567 1.518 1.567 cp_fm_cholesky_invert 11 10.9 1.395 1.399 1.395 1.399 wfi_extrapolate 11 7.9 0.001 0.001 1.388 1.388 potential_pw2rs 119 12.3 0.014 0.014 1.333 1.334 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.299 1.319 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.301 1.301 cp_fm_diag_elpa_base 50 14.0 1.156 1.207 1.299 1.299 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.258 1.263 hybrid_alltoall_any 4725 16.4 0.091 0.151 1.025 1.210 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.144 1.165 mp_alltoall_d11v 2130 13.8 1.141 1.159 1.141 1.159 make_images_data 4572 15.5 0.046 0.049 0.954 1.141 acc_transpose_blocks_sync 27432 16.5 1.092 1.135 1.092 1.135 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.992 1.033 qs_create_task_list 11 7.9 0.018 0.037 0.961 0.974 generate_qs_task_list 11 8.9 0.367 0.387 0.943 0.971 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.918 0.931 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.398000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=731.272727, yerr=17.540037 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.669312E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.052 80.856 80.857 qs_mol_dyn_low 1 2.0 0.003 0.003 80.558 80.567 qs_forces 11 3.9 0.003 0.003 80.460 80.461 qs_energies 11 4.9 0.002 0.007 77.624 77.636 scf_env_do_scf 11 5.9 0.000 0.001 68.886 68.889 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 63.467 63.470 dbcsr_multiply_generic 2055 12.4 0.104 0.108 50.533 50.876 qs_scf_new_mos 99 7.5 0.000 0.001 46.698 46.810 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.697 46.810 ot_scf_mini 99 9.5 0.002 0.002 44.347 44.458 multiply_cannon 2055 13.4 0.184 0.190 42.345 43.084 velocity_verlet 10 3.0 0.005 0.011 42.905 42.905 multiply_cannon_loop 2055 14.4 1.815 1.856 41.400 42.153 ot_mini 99 10.5 0.001 0.001 26.399 26.499 qs_ot_get_derivative 99 11.5 0.001 0.001 19.520 19.629 multiply_cannon_multrec 49320 15.4 11.347 12.092 17.414 18.043 rebuild_ks_matrix 110 8.3 0.000 0.000 14.353 14.472 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 14.352 14.471 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.580 12.692 mp_waitall_1 220248 16.4 10.470 11.448 10.470 11.448 multiply_cannon_sync_h2d 49320 15.4 9.575 10.248 9.575 10.248 qs_ot_get_p 110 10.4 0.001 0.002 9.455 9.624 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.553 8.084 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.232 7.819 apply_single 110 13.6 0.000 0.000 7.231 7.819 multiply_cannon_metrocomm3 49320 15.4 0.082 0.085 6.425 7.542 sum_up_and_integrate 110 10.3 0.002 0.003 6.898 6.910 integrate_v_rspace 110 11.3 0.003 0.003 6.873 6.891 init_scf_run 11 5.9 0.000 0.001 6.696 6.697 scf_env_initial_rho_setup 11 6.9 0.001 0.004 6.696 6.697 ot_diis_step 99 11.5 0.006 0.006 6.610 6.610 qs_ot_p2m_diag 48 11.0 0.013 0.019 6.298 6.320 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.157 6.309 calculate_rho_elec 110 8.6 0.020 0.024 6.157 6.308 dbcsr_mm_accdrv_process 87628 16.1 3.121 3.212 5.936 6.217 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.518 5.518 init_scf_loop 11 6.9 0.001 0.006 5.394 5.414 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.201 5.271 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.968 4.990 cp_fm_diag_elpa_base 48 14.0 4.956 4.977 4.967 4.989 mp_sum_l 6594 12.7 3.940 4.671 3.940 4.671 wfi_extrapolate 11 7.9 0.001 0.001 4.006 4.006 make_m2s 4110 13.4 0.061 0.067 3.792 3.916 make_images 4110 14.4 0.178 0.189 3.697 3.825 calculate_dm_sparse 110 9.5 0.001 0.001 3.709 3.815 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.514 3.518 grid_integrate_task_list 110 12.3 3.245 3.407 3.245 3.407 density_rs2pw 110 9.6 0.004 0.005 3.213 3.390 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.222 3.287 prepare_preconditioner 11 7.9 0.000 0.000 3.224 3.241 make_preconditioner 11 8.9 0.002 0.013 3.224 3.241 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.172 3.213 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 2.283 3.210 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.018 3.068 pw_transfer 1331 11.6 0.055 0.065 2.976 3.051 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.888 2.964 acc_transpose_blocks 49320 15.4 0.215 0.230 2.588 2.645 calculate_first_density_matrix 1 7.0 0.001 0.006 2.594 2.597 fft_wrap_pw1pw2_140 451 13.1 0.309 0.330 2.484 2.572 jit_kernel_multiply 13 15.9 2.530 2.544 2.530 2.544 potential_pw2rs 110 12.3 0.005 0.006 2.514 2.529 mp_alltoall_d11v 2046 13.8 2.018 2.472 2.018 2.472 grid_collocate_task_list 110 9.6 2.155 2.253 2.155 2.253 fft3d_ps 1111 14.6 0.796 0.885 2.180 2.242 mp_waitany 14300 13.8 1.812 2.061 1.812 2.061 transfer_rs2pw 451 10.6 0.005 0.006 1.814 1.973 mp_sum_d 3889 11.9 1.443 1.962 1.443 1.962 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.902 1.923 make_images_data 4110 15.4 0.043 0.048 1.720 1.842 hybrid_alltoall_any 4261 16.3 0.084 0.483 1.514 1.799 cp_fm_cholesky_invert 11 10.9 1.740 1.744 1.740 1.744 transfer_pw2rs 451 13.1 0.006 0.007 1.706 1.712 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.637 1.657 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=80.857000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.272727, yerr=3.077511 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 589.680640E+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 3058147. 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 0.009 0.027 67.955 67.957 qs_mol_dyn_low 1 2.0 0.003 0.003 67.743 67.754 qs_forces 11 3.9 0.006 0.008 67.675 67.676 qs_energies 11 4.9 0.001 0.001 64.353 64.359 scf_env_do_scf 11 5.9 0.000 0.001 55.967 55.970 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.548 48.549 dbcsr_multiply_generic 2055 12.4 0.115 0.119 37.717 37.878 velocity_verlet 10 3.0 0.001 0.002 35.641 35.642 qs_scf_new_mos 99 7.5 0.001 0.001 32.892 33.009 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.892 33.008 multiply_cannon 2055 13.4 0.224 0.242 31.216 32.381 ot_scf_mini 99 9.5 0.003 0.003 31.228 31.355 multiply_cannon_loop 2055 14.4 1.171 1.202 29.967 30.828 ot_mini 99 10.5 0.001 0.001 18.269 18.403 multiply_cannon_multrec 24660 15.4 6.929 8.721 14.026 15.600 rebuild_ks_matrix 110 8.3 0.000 0.000 13.413 13.546 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.413 13.546 qs_ot_get_derivative 99 11.5 0.001 0.001 12.511 12.641 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.784 11.898 mp_waitall_1 176588 16.5 7.421 10.256 7.421 10.256 multiply_cannon_metrocomm3 24660 15.4 0.073 0.074 5.112 8.168 multiply_cannon_sync_h2d 24660 15.4 6.316 7.480 6.316 7.480 init_scf_loop 11 6.9 0.000 0.000 7.385 7.386 dbcsr_mm_accdrv_process 52282 16.1 5.592 6.424 6.927 7.257 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.438 7.190 apply_single 110 13.6 0.000 0.001 6.437 7.190 qs_ot_get_p 110 10.4 0.001 0.002 6.339 6.482 sum_up_and_integrate 110 10.3 0.001 0.002 6.300 6.317 integrate_v_rspace 110 11.3 0.002 0.003 6.273 6.289 init_scf_run 11 5.9 0.000 0.001 6.025 6.026 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.025 6.026 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.788 5.798 calculate_rho_elec 110 8.6 0.039 0.047 5.787 5.797 ot_diis_step 99 11.5 0.010 0.010 5.711 5.712 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.705 5.433 prepare_preconditioner 11 7.9 0.000 0.000 5.393 5.411 make_preconditioner 11 8.9 0.000 0.000 5.393 5.411 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.953 5.103 make_m2s 4110 13.4 0.056 0.059 4.144 4.584 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.457 4.477 make_images 4110 14.4 0.406 0.468 4.032 4.469 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.994 3.994 pw_transfer 1331 11.6 0.066 0.074 3.491 3.625 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.385 3.522 wfi_extrapolate 11 7.9 0.001 0.001 3.470 3.470 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.437 3.449 cp_fm_diag_elpa_base 48 14.0 3.390 3.406 3.434 3.445 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.270 3.343 grid_integrate_task_list 110 12.3 3.171 3.334 3.171 3.334 density_rs2pw 110 9.6 0.004 0.004 3.047 3.274 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.210 3.211 fft_wrap_pw1pw2_140 451 13.1 0.358 0.377 2.913 3.052 calculate_dm_sparse 110 9.5 0.001 0.001 2.965 2.995 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.879 2.920 hybrid_alltoall_any 4261 16.3 0.104 0.452 1.968 2.733 make_images_data 4110 15.4 0.048 0.053 2.250 2.727 fft3d_ps 1111 14.6 1.116 1.344 2.476 2.624 cp_fm_cholesky_invert 11 10.9 2.488 2.495 2.488 2.495 calculate_first_density_matrix 1 7.0 0.000 0.000 2.459 2.461 mp_sum_l 6594 12.7 1.725 2.316 1.725 2.316 grid_collocate_task_list 110 9.6 2.170 2.312 2.170 2.312 mp_alltoall_d11v 2046 13.8 1.714 2.200 1.714 2.200 potential_pw2rs 110 12.3 0.008 0.009 2.177 2.183 acc_transpose_blocks 24660 15.4 0.118 0.122 1.987 2.019 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.956 1.979 jit_kernel_multiply 10 16.4 0.969 1.897 0.969 1.897 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.859 1.863 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.783 1.798 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.775 1.733 transfer_rs2pw 451 10.6 0.006 0.007 1.383 1.632 mp_irecv_dv 57340 16.2 0.649 1.615 0.649 1.615 mp_allgather_i34 2055 14.4 0.557 1.552 0.557 1.552 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.534 1.545 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.387 1.493 mp_sum_d 3889 11.9 0.893 1.426 0.893 1.426 mp_waitany 10164 13.8 1.152 1.417 1.152 1.417 dbcsr_complete_redistribute 325 12.2 0.237 0.309 1.140 1.402 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=67.957000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.727273, yerr=4.769046 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 667.697152E+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 0.010 0.029 59.152 59.154 qs_mol_dyn_low 1 2.0 0.003 0.003 58.768 58.780 qs_forces 11 3.9 0.003 0.003 58.703 58.704 qs_energies 11 4.9 0.001 0.002 55.523 55.526 scf_env_do_scf 11 5.9 0.001 0.001 47.712 47.712 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 39.355 39.356 velocity_verlet 10 3.0 0.002 0.002 32.146 32.148 dbcsr_multiply_generic 2055 12.4 0.111 0.114 28.543 28.794 qs_scf_new_mos 99 7.5 0.001 0.001 24.876 24.973 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.876 24.972 ot_scf_mini 99 9.5 0.002 0.003 23.630 23.742 multiply_cannon 2055 13.4 0.213 0.223 22.429 23.690 multiply_cannon_loop 2055 14.4 0.819 0.847 21.298 22.463 ot_mini 99 10.5 0.001 0.001 13.568 13.682 rebuild_ks_matrix 110 8.3 0.000 0.000 12.040 12.172 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.018 12.040 12.172 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.590 10.711 multiply_cannon_multrec 16440 15.4 3.623 4.507 9.843 10.612 mp_waitall_1 139946 16.5 6.807 9.851 6.807 9.851 qs_ot_get_derivative 99 11.5 0.001 0.001 9.150 9.262 init_scf_loop 11 6.9 0.000 0.000 8.325 8.327 multiply_cannon_metrocomm3 16440 15.4 0.045 0.047 4.334 7.219 prepare_preconditioner 11 7.9 0.000 0.000 6.583 6.601 make_preconditioner 11 8.9 0.000 0.000 6.583 6.601 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.922 6.277 dbcsr_mm_accdrv_process 34862 16.1 5.374 5.768 6.064 6.211 sum_up_and_integrate 110 10.3 0.001 0.002 6.197 6.211 integrate_v_rspace 110 11.3 0.003 0.003 6.171 6.185 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.661 5.671 calculate_rho_elec 110 8.6 0.059 0.059 5.660 5.670 init_scf_run 11 5.9 0.000 0.001 5.423 5.423 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.423 5.423 qs_ot_get_p 110 10.4 0.001 0.002 5.238 5.388 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.909 5.359 apply_single 110 13.6 0.000 0.000 4.909 5.359 make_m2s 4110 13.4 0.049 0.051 4.103 4.465 ot_diis_step 99 11.5 0.011 0.011 4.387 4.388 make_images 4110 14.4 0.398 0.523 3.987 4.347 multiply_cannon_sync_h2d 16440 15.4 3.247 3.894 3.247 3.894 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.143 3.751 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.649 3.653 pw_transfer 1331 11.6 0.065 0.070 3.560 3.568 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.453 3.465 grid_integrate_task_list 110 12.3 3.207 3.351 3.207 3.351 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.283 3.283 density_rs2pw 110 9.6 0.004 0.004 2.917 3.098 fft_wrap_pw1pw2_140 451 13.1 0.461 0.471 3.005 3.017 wfi_extrapolate 11 7.9 0.001 0.001 2.947 2.947 hybrid_alltoall_any 4261 16.3 0.107 0.377 2.087 2.786 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.782 2.784 make_images_data 4110 15.4 0.046 0.051 2.315 2.776 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.765 2.776 cp_fm_diag_elpa_base 48 14.0 2.701 2.734 2.764 2.774 cp_fm_cholesky_invert 11 10.9 2.554 2.561 2.554 2.561 calculate_dm_sparse 110 9.5 0.001 0.001 2.517 2.549 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.449 2.508 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.401 2.461 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.890 2.399 calculate_first_density_matrix 1 7.0 0.000 0.000 2.385 2.387 fft3d_ps 1111 14.6 1.119 1.131 2.369 2.385 grid_collocate_task_list 110 9.6 2.218 2.345 2.218 2.345 mp_irecv_dv 48980 15.7 0.815 2.268 0.815 2.268 potential_pw2rs 110 12.3 0.011 0.011 2.058 2.063 mp_sum_l 6594 12.7 1.382 2.040 1.382 2.040 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.964 1.965 mp_alltoall_d11v 2046 13.8 1.731 1.962 1.731 1.962 dbcsr_complete_redistribute 325 12.2 0.330 0.373 1.388 1.833 cp_fm_upper_to_full 70 14.2 1.390 1.753 1.390 1.753 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.640 1.660 acc_transpose_blocks 16440 15.4 0.080 0.082 1.590 1.642 cp_fm_cholesky_decompose 22 10.9 1.523 1.546 1.523 1.546 mp_allgather_i34 2055 14.4 0.440 1.501 0.440 1.501 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.375 1.487 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.454 1.466 transfer_rs2pw 451 10.6 0.005 0.005 1.232 1.427 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.966 1.410 mp_waitany 17072 13.8 1.122 1.337 1.122 1.337 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.261 1.269 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.154000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=632.090909, yerr=8.532873 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 744.771584E+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 0.015 0.033 65.325 65.326 qs_mol_dyn_low 1 2.0 0.003 0.004 65.064 65.073 qs_forces 11 3.9 0.003 0.006 64.894 64.894 qs_energies 11 4.9 0.002 0.008 61.518 61.522 scf_env_do_scf 11 5.9 0.001 0.001 53.185 53.187 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.714 41.714 velocity_verlet 10 3.0 0.002 0.008 37.076 37.083 dbcsr_multiply_generic 2055 12.4 0.119 0.123 30.584 30.762 qs_scf_new_mos 99 7.5 0.001 0.001 27.184 27.294 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.184 27.294 ot_scf_mini 99 9.5 0.003 0.003 25.500 25.629 multiply_cannon 2055 13.4 0.237 0.252 23.478 24.563 multiply_cannon_loop 2055 14.4 1.439 1.496 22.150 22.682 ot_mini 99 10.5 0.001 0.001 14.721 14.849 multiply_cannon_multrec 24660 15.4 4.066 6.677 13.011 14.247 rebuild_ks_matrix 110 8.3 0.000 0.000 11.929 12.053 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.019 11.929 12.052 init_scf_loop 11 6.9 0.001 0.004 11.433 11.434 qs_ot_get_derivative 99 11.5 0.001 0.001 10.562 10.683 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.542 10.653 dbcsr_mm_accdrv_process 52304 16.0 7.758 9.193 8.786 9.710 prepare_preconditioner 11 7.9 0.000 0.000 9.688 9.706 make_preconditioner 11 8.9 0.000 0.002 9.688 9.706 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.948 9.408 sum_up_and_integrate 110 10.3 0.002 0.003 6.174 6.187 integrate_v_rspace 110 11.3 0.003 0.004 6.147 6.161 mp_waitall_1 121746 16.5 4.159 5.972 4.159 5.972 qs_ot_get_p 110 10.4 0.001 0.001 5.647 5.803 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.792 5.799 calculate_rho_elec 110 8.6 0.077 0.081 5.791 5.798 make_m2s 4110 13.4 0.059 0.061 5.475 5.770 make_images 4110 14.4 0.577 0.702 5.334 5.624 init_scf_run 11 5.9 0.000 0.001 5.546 5.547 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.546 5.547 cp_fm_upper_to_full 70 14.2 3.348 4.853 3.348 4.853 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.036 4.164 apply_single 110 13.6 0.000 0.000 4.035 4.164 ot_diis_step 99 11.5 0.011 0.011 4.120 4.120 qs_ot_p2m_diag 48 11.0 0.055 0.066 3.809 3.825 pw_transfer 1331 11.6 0.065 0.072 3.715 3.748 dbcsr_complete_redistribute 325 12.2 0.415 0.467 2.586 3.738 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.609 3.644 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.507 3.564 grid_integrate_task_list 110 12.3 3.284 3.482 3.284 3.482 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.343 3.411 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.245 3.245 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.099 3.239 fft_wrap_pw1pw2_140 451 13.1 0.530 0.545 3.120 3.159 density_rs2pw 110 9.6 0.004 0.004 2.910 3.077 hybrid_alltoall_any 4261 16.3 0.122 0.457 2.331 3.053 make_images_data 4110 15.4 0.048 0.052 2.748 3.052 calculate_dm_sparse 110 9.5 0.001 0.001 2.977 3.015 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.368 2.998 wfi_extrapolate 11 7.9 0.001 0.001 2.968 2.968 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.731 2.855 mp_alltoall_i22 605 13.7 1.668 2.842 1.668 2.842 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.828 2.831 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.728 2.739 cp_fm_diag_elpa_base 48 14.0 2.583 2.640 2.725 2.737 cp_fm_cholesky_invert 11 10.9 2.562 2.570 2.562 2.570 acc_transpose_blocks 24660 15.4 0.114 0.116 2.456 2.565 multiply_cannon_sync_h2d 24660 15.4 2.408 2.531 2.408 2.531 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.455 2.498 calculate_first_density_matrix 1 7.0 0.000 0.003 2.481 2.485 fft3d_ps 1111 14.6 1.150 1.189 2.435 2.453 grid_collocate_task_list 110 9.6 2.268 2.417 2.268 2.417 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.291 2.292 mp_alltoall_d11v 2046 13.8 1.740 2.074 1.740 2.074 potential_pw2rs 110 12.3 0.012 0.013 1.990 1.993 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.795 1.829 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.613 1.706 cp_fm_cholesky_decompose 22 10.9 1.614 1.668 1.614 1.668 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.629 1.640 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.534 1.549 acc_transpose_blocks_sync 73980 16.4 1.443 1.543 1.443 1.543 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.831 1.522 mp_sum_l 6594 12.7 0.939 1.477 0.939 1.477 jit_kernel_multiply 8 15.6 0.681 1.442 0.681 1.442 mp_irecv_dv 62702 16.1 0.729 1.440 0.729 1.440 mp_allgather_i34 2055 14.4 0.409 1.347 0.409 1.347 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.229 1.339 transfer_rs2pw 451 10.6 0.005 0.006 1.182 1.317 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.326000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=702.545455, yerr=12.921889 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 866.463744E+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 0.013 0.036 54.649 54.653 qs_mol_dyn_low 1 2.0 0.003 0.003 54.407 54.418 qs_forces 11 3.9 0.003 0.003 54.341 54.342 qs_energies 11 4.9 0.001 0.002 50.706 50.710 scf_env_do_scf 11 5.9 0.001 0.001 42.519 42.519 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.881 34.882 velocity_verlet 10 3.0 0.005 0.005 30.923 30.926 dbcsr_multiply_generic 2055 12.4 0.109 0.110 23.087 23.239 qs_scf_new_mos 99 7.5 0.001 0.001 20.371 20.420 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.370 20.420 ot_scf_mini 99 9.5 0.002 0.002 19.121 19.156 multiply_cannon 2055 13.4 0.238 0.252 17.586 19.034 multiply_cannon_loop 2055 14.4 0.606 0.630 16.318 16.689 rebuild_ks_matrix 110 8.3 0.000 0.000 11.612 11.648 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.612 11.647 ot_mini 99 10.5 0.001 0.001 10.608 10.636 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.326 10.360 multiply_cannon_multrec 8220 15.4 3.195 4.414 7.608 8.578 init_scf_loop 11 6.9 0.000 0.000 7.593 7.594 mp_waitall_1 103326 16.6 5.757 7.323 5.757 7.323 qs_ot_get_derivative 99 11.5 0.001 0.001 6.935 6.968 sum_up_and_integrate 110 10.3 0.001 0.002 6.176 6.188 integrate_v_rspace 110 11.3 0.003 0.003 6.149 6.162 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.955 5.966 calculate_rho_elec 110 8.6 0.113 0.113 5.955 5.965 prepare_preconditioner 11 7.9 0.000 0.000 5.955 5.961 make_preconditioner 11 8.9 0.000 0.000 5.955 5.960 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.547 5.622 dbcsr_mm_accdrv_process 17442 15.9 3.161 4.106 4.273 5.197 init_scf_run 11 5.9 0.000 0.001 5.080 5.080 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.079 5.080 qs_ot_get_p 110 10.4 0.002 0.002 4.568 4.590 make_m2s 4110 13.4 0.037 0.039 4.189 4.477 make_images 4110 14.4 0.647 0.707 4.058 4.348 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.924 4.304 pw_transfer 1331 11.6 0.066 0.070 4.046 4.056 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.938 3.952 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.707 3.769 apply_single 110 13.6 0.000 0.000 3.707 3.769 ot_diis_step 99 11.5 0.012 0.012 3.651 3.651 grid_integrate_task_list 110 12.3 3.380 3.475 3.380 3.475 fft_wrap_pw1pw2_140 451 13.1 0.718 0.732 3.428 3.443 qs_ot_p2m_diag 48 11.0 0.081 0.085 3.224 3.228 density_rs2pw 110 9.6 0.004 0.004 2.984 3.151 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.889 2.889 cp_fm_cholesky_invert 11 10.9 2.821 2.825 2.821 2.825 wfi_extrapolate 11 7.9 0.001 0.001 2.699 2.699 hybrid_alltoall_any 4261 16.3 0.199 0.850 2.232 2.688 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.659 2.659 make_images_data 4110 15.4 0.041 0.047 2.253 2.630 grid_collocate_task_list 110 9.6 2.365 2.551 2.365 2.551 calculate_dm_sparse 110 9.5 0.001 0.001 2.477 2.516 fft3d_ps 1111 14.6 1.289 1.297 2.483 2.501 multiply_cannon_sync_h2d 8220 15.4 2.367 2.499 2.367 2.499 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.492 2.493 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.391 2.399 cp_fm_diag_elpa_base 48 14.0 2.337 2.359 2.390 2.397 calculate_first_density_matrix 1 7.0 0.000 0.000 2.283 2.286 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.123 2.143 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.000 2.036 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.783 1.989 potential_pw2rs 110 12.3 0.015 0.015 1.980 1.985 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.899 1.920 mp_alltoall_d11v 2046 13.8 1.614 1.763 1.614 1.763 cp_fm_cholesky_decompose 22 10.9 1.677 1.701 1.677 1.701 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.612 1.616 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.491 1.602 mp_allgather_i34 2055 14.4 0.439 1.528 0.439 1.528 dbcsr_complete_redistribute 325 12.2 0.565 0.586 1.410 1.505 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.429 1.439 acc_transpose_blocks 8220 15.4 0.041 0.043 1.360 1.397 qs_create_task_list 11 7.9 0.001 0.001 1.205 1.304 generate_qs_task_list 11 8.9 0.374 0.442 1.204 1.303 transfer_rs2pw 451 10.6 0.005 0.005 1.087 1.277 mp_waitany 9240 13.8 1.031 1.233 1.031 1.233 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.158 1.183 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.469 1.183 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.769 1.154 mp_irecv_dv 24056 15.7 0.443 1.137 0.443 1.137 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.653000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=813.363636, yerr=14.710990 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 1.340420E+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 0.018 0.039 86.646 86.647 qs_mol_dyn_low 1 2.0 0.003 0.003 86.307 86.320 qs_forces 11 3.9 0.003 0.003 86.235 86.235 qs_energies 11 4.9 0.001 0.001 82.081 82.083 scf_env_do_scf 11 5.9 0.001 0.001 72.129 72.129 velocity_verlet 10 3.0 0.002 0.002 55.395 55.402 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.232 44.233 dbcsr_multiply_generic 2055 12.4 0.123 0.125 29.224 29.314 init_scf_loop 11 6.9 0.000 0.000 27.828 27.828 qs_scf_new_mos 99 7.5 0.001 0.001 26.416 26.449 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.415 26.448 prepare_preconditioner 11 7.9 0.000 0.000 25.810 25.818 make_preconditioner 11 8.9 0.000 0.000 25.810 25.818 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.056 25.295 ot_scf_mini 99 9.5 0.002 0.002 24.634 24.662 multiply_cannon 2055 13.4 0.336 0.369 22.301 23.040 multiply_cannon_loop 2055 14.4 0.833 0.846 20.534 20.835 cp_fm_upper_to_full 70 14.2 12.637 18.122 12.637 18.122 ot_mini 99 10.5 0.001 0.001 13.943 13.971 rebuild_ks_matrix 110 8.3 0.000 0.001 13.904 13.937 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.903 13.937 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.605 12.636 dbcsr_complete_redistribute 325 12.2 1.003 1.021 7.282 10.513 multiply_cannon_multrec 8220 15.4 4.035 4.183 9.747 9.873 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.287 9.514 qs_ot_get_derivative 99 11.5 0.001 0.001 9.387 9.416 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.739 8.943 mp_alltoall_i22 605 13.7 5.361 8.570 5.361 8.570 mp_waitall_1 84994 16.7 7.311 8.167 7.311 8.167 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.427 7.462 calculate_rho_elec 110 8.6 0.222 0.223 7.426 7.462 sum_up_and_integrate 110 10.3 0.002 0.002 6.877 6.891 integrate_v_rspace 110 11.3 0.004 0.004 6.849 6.864 dbcsr_mm_accdrv_process 11614 15.7 3.883 4.173 5.563 5.781 pw_transfer 1331 11.6 0.074 0.075 5.765 5.770 make_m2s 4110 13.4 0.043 0.043 5.239 5.729 init_scf_run 11 5.9 0.000 0.001 5.670 5.670 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.669 5.669 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.648 5.653 make_images 4110 14.4 0.885 0.938 5.050 5.539 qs_ot_get_p 110 10.4 0.001 0.002 5.419 5.445 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.708 5.178 apply_single 110 13.6 0.000 0.000 4.708 5.177 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.755 5.049 cp_fm_cholesky_invert 11 10.9 4.982 4.986 4.982 4.986 fft_wrap_pw1pw2_140 451 13.1 1.368 1.372 4.972 4.978 ot_diis_step 99 11.5 0.015 0.015 4.540 4.540 density_rs2pw 110 9.6 0.004 0.004 3.899 3.912 qs_ot_p2m_diag 48 11.0 0.151 0.157 3.888 3.896 grid_integrate_task_list 110 12.3 3.679 3.753 3.679 3.753 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.678 3.678 hybrid_alltoall_any 4261 16.3 0.264 0.566 2.767 3.492 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.992 3.449 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.369 3.370 fft3d_ps 1111 14.6 1.879 1.888 3.352 3.359 make_images_data 4110 15.4 0.045 0.048 2.732 3.332 wfi_extrapolate 11 7.9 0.001 0.001 3.313 3.313 calculate_dm_sparse 110 9.5 0.001 0.001 3.143 3.166 multiply_cannon_sync_h2d 8220 15.4 3.116 3.136 3.116 3.136 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.849 2.851 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.797 2.798 cp_fm_diag_elpa_base 48 14.0 2.263 2.458 2.795 2.795 grid_collocate_task_list 110 9.6 2.669 2.684 2.669 2.684 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.613 2.628 potential_pw2rs 110 12.3 0.021 0.021 2.418 2.420 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.361 2.373 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.197 2.251 calculate_first_density_matrix 1 7.0 0.000 0.000 2.239 2.240 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.103 2.188 mp_alltoall_d11v 2046 13.8 1.986 2.059 1.986 2.059 cp_fm_cholesky_decompose 22 10.9 1.922 1.945 1.922 1.945 qs_create_task_list 11 7.9 0.001 0.001 1.877 1.924 generate_qs_task_list 11 8.9 0.732 0.786 1.877 1.923 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.896 1.899 acc_transpose_blocks 8220 15.4 0.042 0.043 1.725 1.753 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.717 1.748 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.647000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1223.181818, yerr=46.504776 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.066624E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 3243333. 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 0.010 0.031 207.781 207.783 qs_mol_dyn_low 1 2.0 0.003 0.003 207.355 207.370 qs_forces 11 3.9 0.005 0.005 207.266 207.267 qs_energies 11 4.9 0.001 0.002 201.662 201.678 scf_env_do_scf 11 5.9 0.001 0.003 185.221 185.226 scf_env_do_scf_inner_loop 117 6.6 0.002 0.009 164.720 164.722 dbcsr_multiply_generic 2507 12.6 0.178 0.181 126.053 126.632 qs_scf_new_mos 117 7.6 0.001 0.001 125.257 125.575 qs_scf_loop_do_ot 117 8.6 0.001 0.001 125.256 125.574 velocity_verlet 10 3.0 0.001 0.002 124.076 124.078 ot_scf_mini 117 9.6 0.003 0.003 118.519 118.828 multiply_cannon 2507 13.6 0.236 0.245 101.685 103.233 multiply_cannon_loop 2507 14.6 2.432 2.485 99.516 100.947 ot_mini 117 10.6 0.001 0.001 66.449 66.747 multiply_cannon_multrec 60168 15.6 31.642 33.609 41.513 43.429 qs_ot_get_derivative 117 11.6 0.001 0.001 41.748 42.026 rebuild_ks_matrix 128 8.3 0.001 0.001 33.567 33.908 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.566 33.907 mp_waitall_1 267128 16.5 29.198 32.825 29.198 32.825 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.148 30.467 qs_ot_get_p 128 10.4 0.001 0.001 29.773 30.132 multiply_cannon_sync_h2d 60168 15.6 26.091 27.780 26.091 27.780 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.271 25.080 apply_single 128 13.6 0.001 0.001 24.271 25.080 ot_diis_step 117 11.6 0.008 0.008 24.449 24.450 qs_ot_p2m_diag 83 11.4 0.079 0.092 23.073 23.142 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.204 20.441 init_scf_loop 11 6.9 0.000 0.000 20.427 20.428 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.238 20.239 multiply_cannon_metrocomm3 60168 15.6 0.116 0.122 16.269 18.324 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.252 17.289 cp_fm_diag_elpa_base 83 14.4 17.185 17.224 17.248 17.286 prepare_preconditioner 11 7.9 0.000 0.000 15.805 15.857 make_preconditioner 11 8.9 0.000 0.000 15.805 15.857 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.036 15.213 make_m2s 5014 13.6 0.104 0.113 13.771 14.200 sum_up_and_integrate 128 10.3 0.002 0.004 14.026 14.046 make_images 5014 14.6 0.400 0.417 13.586 14.021 integrate_v_rspace 128 11.3 0.003 0.004 13.967 13.988 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.182 13.319 calculate_rho_elec 128 8.7 0.045 0.064 13.182 13.319 init_scf_run 11 5.9 0.000 0.001 12.324 12.324 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.323 12.324 mp_sum_l 7950 12.9 9.342 10.500 9.342 10.500 dbcsr_mm_accdrv_process 124484 16.2 4.826 5.000 9.452 10.006 cp_fm_cholesky_invert 11 10.9 9.083 9.090 9.083 9.090 wfi_extrapolate 11 7.9 0.001 0.001 9.030 9.030 calculate_dm_sparse 128 9.5 0.001 0.001 8.559 8.681 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.242 8.350 multiply_cannon_metrocomm1 60168 15.6 0.094 0.096 6.458 8.288 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.159 8.272 pw_transfer 1547 11.6 0.073 0.090 7.603 7.789 make_images_data 5014 15.6 0.066 0.072 6.718 7.695 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.401 7.587 grid_integrate_task_list 128 12.3 7.057 7.439 7.057 7.439 density_rs2pw 128 9.7 0.006 0.007 6.764 7.370 hybrid_alltoall_any 5200 16.5 0.293 2.273 5.892 7.223 fft_wrap_pw1pw2_140 523 13.2 0.860 0.906 6.492 6.688 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.671 6.683 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.784 5.912 fft3d_ps 1291 14.7 2.195 2.857 5.405 5.737 mp_alltoall_d11v 2415 14.1 4.369 5.528 4.369 5.528 grid_collocate_task_list 128 9.7 4.846 5.287 4.846 5.287 potential_pw2rs 128 12.3 0.009 0.010 4.531 4.557 cp_fm_cholesky_decompose 22 10.9 4.544 4.556 4.544 4.556 mp_sum_d 4465 12.1 3.437 4.349 3.437 4.349 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.783000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.090909, yerr=6.052115 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 848.769024E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58142. MP_Allreduce 11096 960. MP_Sync 86 MP_Alltoall 1955 5721178. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 187.215 187.217 qs_mol_dyn_low 1 2.0 0.003 0.006 186.827 186.844 qs_forces 11 3.9 0.004 0.005 186.738 186.741 qs_energies 11 4.9 0.001 0.002 179.986 179.997 scf_env_do_scf 11 5.9 0.005 0.039 163.717 163.727 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 131.065 131.068 velocity_verlet 10 3.0 0.001 0.002 117.379 117.381 dbcsr_multiply_generic 2485 12.5 0.186 0.192 96.685 97.944 qs_scf_new_mos 116 7.6 0.001 0.001 93.031 93.517 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.030 93.516 ot_scf_mini 116 9.6 0.004 0.004 88.257 88.882 multiply_cannon 2485 13.5 0.499 0.568 76.622 81.620 multiply_cannon_loop 2485 14.5 1.564 1.633 73.231 75.873 ot_mini 116 10.6 0.001 0.001 49.777 50.314 mp_waitall_1 212858 16.6 23.821 38.112 23.821 38.112 multiply_cannon_multrec 29820 15.5 20.705 24.802 31.288 35.680 rebuild_ks_matrix 127 8.3 0.001 0.001 32.108 32.787 qs_ks_build_kohn_sham_matrix 127 9.3 0.031 0.136 32.107 32.786 init_scf_loop 11 6.9 0.000 0.000 32.557 32.563 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.852 29.469 qs_ot_get_derivative 116 11.6 0.001 0.002 28.069 28.681 prepare_preconditioner 11 7.9 0.000 0.000 28.060 28.135 make_preconditioner 11 8.9 0.000 0.000 28.060 28.135 multiply_cannon_metrocomm3 29820 15.5 0.097 0.102 15.554 27.994 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.763 27.296 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.858 22.939 apply_single 127 13.6 0.001 0.001 21.857 22.938 ot_diis_step 116 11.6 0.014 0.015 21.525 21.527 qs_ot_get_p 127 10.4 0.001 0.002 20.687 21.330 multiply_cannon_sync_h2d 29820 15.5 17.752 20.506 17.752 20.506 cp_fm_cholesky_invert 11 10.9 16.403 16.415 16.403 16.415 qs_ot_p2m_diag 82 11.4 0.186 0.214 15.767 15.804 make_m2s 4970 13.5 0.085 0.090 14.087 15.793 make_images 4970 14.5 1.154 1.366 13.872 15.575 cp_dbcsr_syevd 82 12.4 0.005 0.006 14.578 14.579 sum_up_and_integrate 127 10.3 0.006 0.031 13.850 13.880 integrate_v_rspace 127 11.3 0.003 0.004 13.787 13.821 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.174 13.211 calculate_rho_elec 127 8.7 0.086 0.103 13.173 13.210 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.101 11.530 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.480 11.518 cp_fm_diag_elpa_base 82 14.4 11.230 11.327 11.474 11.507 init_scf_run 11 5.9 0.000 0.001 11.471 11.472 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.470 11.472 multiply_cannon_metrocomm4 27335 15.5 0.098 0.111 3.857 10.925 dbcsr_mm_accdrv_process 61726 16.2 5.596 6.418 10.040 10.577 mp_irecv_dv 68888 16.3 3.657 10.526 3.657 10.526 make_images_data 4970 15.5 0.066 0.077 8.288 10.272 hybrid_alltoall_any 5155 16.4 0.344 1.514 7.081 9.713 pw_transfer 1535 11.6 0.084 0.094 8.499 8.556 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.278 8.338 wfi_extrapolate 11 7.9 0.001 0.001 8.303 8.303 grid_integrate_task_list 127 12.3 7.133 7.408 7.133 7.408 fft_wrap_pw1pw2_140 519 13.2 0.927 0.942 7.325 7.399 density_rs2pw 127 9.7 0.006 0.006 6.950 7.360 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.280 7.056 cp_fm_cholesky_decompose 22 10.9 6.833 6.918 6.833 6.918 calculate_dm_sparse 127 9.5 0.001 0.001 6.455 6.640 mp_sum_l 7884 12.9 4.201 6.166 4.201 6.166 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.101 6.109 fft3d_ps 1281 14.7 2.805 2.975 5.852 5.901 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.224 5.400 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.311 5.393 grid_collocate_task_list 127 9.7 4.960 5.270 4.960 5.270 mp_allgather_i34 2485 14.5 1.920 4.995 1.920 4.995 mp_alltoall_d11v 2401 14.1 4.056 4.588 4.056 4.588 potential_pw2rs 127 12.3 0.015 0.017 4.406 4.417 mp_sum_d 4454 12.1 2.685 4.102 2.685 4.102 dbcsr_complete_redistribute 393 12.7 0.762 0.847 3.065 3.892 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=187.217000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=806.363636, yerr=5.580367 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/17/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 2.906045E+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 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 947.740672E+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 925172558800 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58307. MP_Allreduce 11065 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 0.016 0.035 172.408 172.409 qs_mol_dyn_low 1 2.0 0.003 0.003 172.005 172.020 qs_forces 11 3.9 0.004 0.004 171.912 171.914 qs_energies 11 4.9 0.002 0.008 165.403 165.415 scf_env_do_scf 11 5.9 0.001 0.002 149.978 149.980 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.425 115.426 velocity_verlet 10 3.0 0.002 0.002 110.354 110.356 dbcsr_multiply_generic 2485 12.5 0.186 0.192 81.306 82.555 qs_scf_new_mos 116 7.6 0.001 0.001 78.820 79.160 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.819 79.159 ot_scf_mini 116 9.6 0.004 0.004 74.620 75.043 multiply_cannon 2485 13.5 0.504 0.532 61.891 67.296 multiply_cannon_loop 2485 14.5 1.124 1.189 58.842 61.693 ot_mini 116 10.6 0.001 0.002 42.177 42.586 init_scf_loop 11 6.9 0.001 0.003 34.457 34.458 mp_waitall_1 169034 16.6 24.261 34.442 24.261 34.442 prepare_preconditioner 11 7.9 0.000 0.000 30.460 30.503 make_preconditioner 11 8.9 0.000 0.002 30.459 30.503 rebuild_ks_matrix 127 8.3 0.001 0.001 29.892 30.365 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.021 29.891 30.364 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.107 29.499 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.916 27.351 multiply_cannon_multrec 19880 15.5 13.009 15.634 22.311 25.023 multiply_cannon_metrocomm3 19880 15.5 0.062 0.066 15.150 24.639 qs_ot_get_derivative 116 11.6 0.001 0.002 22.871 23.293 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.416 20.451 apply_single 127 13.6 0.001 0.001 19.415 20.450 ot_diis_step 116 11.6 0.018 0.020 19.196 19.197 qs_ot_get_p 127 10.4 0.002 0.003 17.929 18.428 make_m2s 4970 13.5 0.077 0.084 14.337 15.716 make_images 4970 14.5 1.133 1.229 14.102 15.477 multiply_cannon_sync_h2d 19880 15.5 13.556 15.342 13.556 15.342 cp_fm_cholesky_invert 11 10.9 14.062 14.071 14.062 14.071 sum_up_and_integrate 127 10.3 0.002 0.003 13.753 13.779 integrate_v_rspace 127 11.3 0.003 0.005 13.694 13.721 qs_ot_p2m_diag 82 11.4 0.263 0.274 13.634 13.645 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.377 13.400 calculate_rho_elec 127 8.7 0.128 0.142 13.376 13.400 cp_dbcsr_syevd 82 12.4 0.005 0.007 12.600 12.602 init_scf_run 11 5.9 0.000 0.001 10.410 10.411 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.410 10.410 make_images_data 4970 15.5 0.063 0.074 8.583 10.355 hybrid_alltoall_any 5155 16.4 0.442 2.008 7.401 10.012 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.521 9.544 cp_fm_diag_elpa_base 82 14.4 9.119 9.257 9.517 9.540 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.981 9.284 multiply_cannon_metrocomm4 17395 15.5 0.065 0.073 3.451 9.104 dbcsr_mm_accdrv_process 41158 16.2 5.681 5.983 8.763 8.893 mp_irecv_dv 49801 16.2 3.324 8.849 3.324 8.849 pw_transfer 1535 11.6 0.084 0.102 8.635 8.736 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.413 8.519 grid_integrate_task_list 127 12.3 7.299 7.707 7.299 7.707 fft_wrap_pw1pw2_140 519 13.2 1.055 1.089 7.498 7.615 wfi_extrapolate 11 7.9 0.001 0.001 7.394 7.394 cp_fm_upper_to_full 104 14.8 5.818 7.333 5.818 7.333 cp_fm_cholesky_decompose 22 10.9 7.158 7.213 7.158 7.213 density_rs2pw 127 9.7 0.005 0.006 6.758 7.177 dbcsr_complete_redistribute 393 12.7 1.167 1.197 4.537 6.304 calculate_dm_sparse 127 9.5 0.001 0.001 5.792 5.887 fft3d_ps 1281 14.7 2.800 3.027 5.745 5.842 grid_collocate_task_list 127 9.7 5.118 5.567 5.118 5.567 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.445 5.450 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.634 5.327 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.404 5.166 mp_sum_l 7884 12.9 3.259 4.699 3.259 4.699 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.551 4.682 mp_alltoall_d11v 2401 14.1 4.245 4.662 4.245 4.662 mp_allgather_i34 2485 14.5 1.553 4.610 1.553 4.610 potential_pw2rs 127 12.3 0.019 0.021 4.230 4.238 transfer_fm_to_dbcsr 11 9.9 0.018 0.026 2.334 4.057 mp_alltoall_i22 712 14.1 1.984 3.914 1.984 3.914 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.855 3.887 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.720 3.720 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.469 3.515 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=172.409000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=897.454545, yerr=7.190444 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/18/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 4.387242E+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 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.177084E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57626. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509604. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.034 187.729 187.730 qs_mol_dyn_low 1 2.0 0.003 0.004 187.326 187.339 qs_forces 11 3.9 0.004 0.006 187.217 187.226 qs_energies 11 4.9 0.003 0.013 180.068 180.079 scf_env_do_scf 11 5.9 0.001 0.003 163.422 163.430 velocity_verlet 10 3.0 0.002 0.002 125.113 125.115 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 117.080 117.083 dbcsr_multiply_generic 2529 12.6 0.197 0.204 80.628 81.137 qs_scf_new_mos 118 7.6 0.001 0.001 80.529 80.853 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.529 80.853 ot_scf_mini 118 9.6 0.003 0.004 75.917 76.228 multiply_cannon 2529 13.6 0.561 0.600 56.187 58.830 multiply_cannon_loop 2529 14.6 1.866 1.927 52.438 53.766 init_scf_loop 11 6.9 0.001 0.004 46.219 46.222 ot_mini 118 10.6 0.001 0.001 42.931 43.240 prepare_preconditioner 11 7.9 0.000 0.000 42.121 42.152 make_preconditioner 11 8.9 0.000 0.003 42.121 42.152 make_full_inverse_cholesky 11 9.9 0.011 0.022 35.626 40.815 multiply_cannon_multrec 30348 15.6 13.790 18.714 26.832 31.468 rebuild_ks_matrix 129 8.3 0.001 0.001 29.486 29.750 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.022 29.485 29.749 mp_waitall_1 149172 16.7 17.013 27.430 17.013 27.430 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.590 26.832 qs_ot_get_derivative 118 11.6 0.001 0.002 23.245 23.557 make_m2s 5058 13.6 0.092 0.097 20.150 21.337 make_images 5058 14.6 1.957 2.266 19.841 21.032 ot_diis_step 118 11.6 0.018 0.019 19.559 19.561 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.144 19.554 apply_single 129 13.6 0.001 0.001 19.143 19.554 qs_ot_get_p 129 10.4 0.002 0.002 19.057 19.394 cp_fm_upper_to_full 106 14.8 11.718 17.304 11.718 17.304 cp_fm_cholesky_invert 11 10.9 15.785 15.793 15.785 15.793 qs_ot_p2m_diag 84 11.4 0.347 0.395 14.920 14.974 multiply_cannon_metrocomm3 30348 15.6 0.049 0.052 6.427 14.801 sum_up_and_integrate 129 10.3 0.002 0.004 14.067 14.096 integrate_v_rspace 129 11.3 0.003 0.004 14.007 14.038 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.879 13.911 calculate_rho_elec 129 8.7 0.173 0.189 13.879 13.911 cp_dbcsr_syevd 84 12.4 0.005 0.007 13.497 13.499 dbcsr_mm_accdrv_process 62780 16.2 8.533 9.367 12.601 13.060 dbcsr_complete_redistribute 397 12.7 1.483 1.598 9.184 13.060 make_images_data 5058 15.6 0.066 0.073 10.766 12.605 hybrid_alltoall_any 5245 16.5 0.534 2.229 9.681 11.834 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.816 11.696 multiply_cannon_sync_h2d 30348 15.6 10.432 11.139 10.432 11.139 init_scf_run 11 5.9 0.000 0.001 10.737 10.739 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.737 10.738 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.302 10.319 cp_fm_diag_elpa_base 84 14.4 9.319 9.660 10.295 10.313 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 6.472 10.247 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.800 10.034 mp_alltoall_i22 720 14.1 5.726 9.677 5.726 9.677 pw_transfer 1559 11.6 0.086 0.104 9.254 9.324 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 9.027 9.103 fft_wrap_pw1pw2_140 527 13.2 1.259 1.315 8.054 8.136 grid_integrate_task_list 129 12.3 7.552 7.958 7.552 7.958 wfi_extrapolate 11 7.9 0.001 0.001 7.650 7.650 cp_fm_cholesky_decompose 22 10.9 7.426 7.511 7.426 7.511 multiply_cannon_metrocomm4 25290 15.6 0.080 0.090 2.780 7.374 density_rs2pw 129 9.7 0.005 0.006 6.852 7.223 mp_irecv_dv 76751 16.2 2.628 7.102 2.628 7.102 calculate_dm_sparse 129 9.5 0.001 0.001 6.291 6.381 fft3d_ps 1301 14.7 3.015 3.094 6.053 6.127 grid_collocate_task_list 129 9.7 5.326 5.655 5.326 5.655 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.456 5.503 mp_alltoall_d11v 2429 14.1 4.960 5.397 4.960 5.397 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.430 4.517 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.372 4.459 qs_energies_init_hamiltonians 11 5.9 0.001 0.006 4.426 4.427 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.266 4.329 potential_pw2rs 129 12.3 0.023 0.024 4.278 4.288 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.730000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1102.909091, yerr=25.432458 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 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 11528908111872 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 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.544360E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350290592 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57947. MP_Allreduce 11127 1165. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.029 169.521 169.521 qs_mol_dyn_low 1 2.0 0.003 0.003 169.128 169.141 qs_forces 11 3.9 0.004 0.004 168.825 168.829 qs_energies 11 4.9 0.001 0.002 161.418 161.425 scf_env_do_scf 11 5.9 0.001 0.001 144.368 144.375 velocity_verlet 10 3.0 0.002 0.002 111.633 111.649 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 108.725 108.726 dbcsr_multiply_generic 2507 12.6 0.189 0.194 72.152 72.561 qs_scf_new_mos 117 7.6 0.001 0.001 71.695 71.808 qs_scf_loop_do_ot 117 8.6 0.001 0.001 71.694 71.807 ot_scf_mini 117 9.6 0.004 0.004 67.267 67.381 multiply_cannon 2507 13.6 0.562 0.577 53.579 56.582 multiply_cannon_loop 2507 14.6 0.808 0.840 50.465 51.481 ot_mini 117 10.6 0.001 0.001 37.790 37.894 init_scf_loop 11 6.9 0.000 0.000 35.493 35.494 prepare_preconditioner 11 7.9 0.000 0.000 31.653 31.682 make_preconditioner 11 8.9 0.000 0.000 31.653 31.682 mp_waitall_1 125778 16.7 24.849 31.442 24.849 31.442 make_full_inverse_cholesky 11 9.9 0.013 0.028 29.518 29.782 rebuild_ks_matrix 128 8.3 0.001 0.001 29.137 29.281 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.022 29.137 29.281 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.589 26.722 multiply_cannon_multrec 10028 15.6 10.300 16.057 17.901 22.679 qs_ot_get_derivative 117 11.6 0.002 0.002 20.495 20.600 multiply_cannon_metrocomm3 10028 15.6 0.024 0.025 12.969 20.457 cp_fm_cholesky_invert 11 10.9 18.124 18.130 18.124 18.130 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 17.333 17.508 apply_single 128 13.6 0.001 0.001 17.333 17.508 ot_diis_step 117 11.6 0.020 0.020 17.214 17.214 qs_ot_get_p 128 10.4 0.002 0.002 16.761 16.916 make_m2s 5014 13.6 0.064 0.069 14.625 15.562 make_images 5014 14.6 2.162 2.593 14.316 15.245 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.208 14.258 calculate_rho_elec 128 8.7 0.254 0.265 14.207 14.257 sum_up_and_integrate 128 10.3 0.002 0.003 14.174 14.224 integrate_v_rspace 128 11.3 0.004 0.004 14.114 14.165 qs_ot_p2m_diag 83 11.4 0.496 0.501 12.984 13.001 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.821 11.822 multiply_cannon_sync_h2d 10028 15.6 10.693 11.322 10.693 11.322 init_scf_run 11 5.9 0.000 0.001 10.266 10.266 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.266 10.266 pw_transfer 1547 11.6 0.085 0.095 9.864 9.900 make_images_data 5014 15.6 0.055 0.066 8.385 9.794 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 9.642 9.684 hybrid_alltoall_any 5200 16.5 0.838 3.767 8.203 9.355 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.972 8.985 cp_fm_diag_elpa_base 83 14.4 8.738 8.815 8.969 8.981 fft_wrap_pw1pw2_140 523 13.2 1.669 1.700 8.539 8.580 grid_integrate_task_list 128 12.3 7.788 8.135 7.788 8.135 dbcsr_mm_accdrv_process 20762 16.1 3.415 4.476 7.255 8.110 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.021 8.095 cp_fm_cholesky_decompose 22 10.9 7.848 7.965 7.848 7.965 multiply_cannon_metrocomm1 10028 15.6 0.030 0.031 4.256 7.484 wfi_extrapolate 11 7.9 0.001 0.001 7.294 7.295 density_rs2pw 128 9.7 0.005 0.005 6.936 7.243 calculate_dm_sparse 128 9.5 0.001 0.001 6.072 6.154 fft3d_ps 1291 14.7 3.141 3.250 6.094 6.129 multiply_cannon_metrocomm4 7521 15.6 0.026 0.030 1.870 5.862 grid_collocate_task_list 128 9.7 5.563 5.839 5.563 5.839 mp_irecv_dv 28860 15.9 1.828 5.766 1.828 5.766 dbcsr_complete_redistribute 395 12.7 2.066 2.116 5.224 5.578 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.219 5.220 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.053 5.068 mp_alltoall_d11v 2415 14.1 4.595 4.954 4.595 4.954 mp_allgather_i34 2507 14.6 1.302 4.603 1.302 4.603 potential_pw2rs 128 12.3 0.026 0.026 4.350 4.356 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.000 4.048 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.579 3.915 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.620 3.714 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.651 3.700 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.382 3.692 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.420 3.517 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.450 3.464 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.521000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1447.636364, yerr=22.740723 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 11.785520E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 1979552 0.0% 0.0% 100.0% average stack size 0.0 0.0 3438.3 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 3.179831E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101080 MPI messages size (bytes): total size 1.144901E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.326685E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45568 35366371328 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726591885696 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 58825. MP_Allreduce 11137 1501. MP_Sync 87 MP_Alltoall 1724 36993750. 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 0.022 0.043 293.725 293.727 qs_mol_dyn_low 1 2.0 0.003 0.003 293.180 293.196 qs_forces 11 3.9 0.005 0.005 293.078 293.080 qs_energies 11 4.9 0.002 0.002 284.196 284.200 scf_env_do_scf 11 5.9 0.001 0.001 261.706 261.715 velocity_verlet 10 3.0 0.002 0.002 212.898 212.906 scf_env_do_scf_inner_loop 118 6.6 0.004 0.008 137.040 137.042 init_scf_loop 11 6.9 0.000 0.000 124.388 124.390 prepare_preconditioner 11 7.9 0.000 0.000 119.537 119.560 make_preconditioner 11 8.9 0.000 0.000 119.537 119.560 make_full_inverse_cholesky 11 9.9 0.036 0.039 95.421 116.656 qs_scf_new_mos 118 7.6 0.001 0.001 91.650 91.758 qs_scf_loop_do_ot 118 8.6 0.001 0.001 91.649 91.758 ot_scf_mini 118 9.6 0.004 0.004 86.748 86.804 dbcsr_multiply_generic 2527 12.6 0.221 0.229 83.177 83.720 cp_fm_upper_to_full 105 14.8 52.597 75.693 52.597 75.693 multiply_cannon 2527 13.6 0.689 0.724 59.541 60.890 multiply_cannon_loop 2527 14.6 1.059 1.080 55.445 56.985 ot_mini 118 10.6 0.001 0.001 45.447 45.531 dbcsr_complete_redistribute 395 12.7 3.966 4.027 30.127 43.217 copy_fm_to_dbcsr 209 11.7 0.001 0.002 26.583 39.659 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.072 36.999 rebuild_ks_matrix 129 8.3 0.001 0.001 35.111 35.169 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 35.111 35.168 mp_alltoall_i22 718 14.1 21.846 35.014 21.846 35.014 cp_fm_cholesky_invert 11 10.9 33.520 33.526 33.520 33.526 mp_waitall_1 104500 16.8 28.659 32.722 28.659 32.722 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.585 32.640 qs_ot_get_p 129 10.4 0.002 0.002 26.232 26.363 qs_ot_get_derivative 118 11.6 0.002 0.002 25.132 25.185 qs_ot_p2m_diag 83 11.4 0.880 0.884 22.018 22.050 multiply_cannon_metrocomm3 10108 15.6 0.025 0.025 19.229 20.680 make_m2s 5054 13.6 0.074 0.076 18.895 20.393 ot_diis_step 118 11.6 0.022 0.022 20.269 20.269 cp_dbcsr_syevd 83 12.4 0.006 0.006 20.202 20.204 make_images 5054 14.6 3.083 3.279 18.407 19.903 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.541 19.706 apply_single 129 13.6 0.001 0.001 19.541 19.705 multiply_cannon_multrec 10108 15.6 10.331 12.185 18.168 18.267 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.525 17.535 calculate_rho_elec 129 8.7 0.482 0.483 17.524 17.534 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.993 16.994 cp_fm_diag_elpa_base 83 14.4 12.593 14.212 16.989 16.990 sum_up_and_integrate 129 10.3 0.002 0.002 16.614 16.702 integrate_v_rspace 129 11.3 0.004 0.005 16.551 16.639 multiply_cannon_sync_h2d 10108 15.6 14.416 14.462 14.416 14.462 pw_transfer 1559 11.6 0.096 0.097 13.360 13.372 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 13.121 13.131 hybrid_alltoall_any 5240 16.5 1.327 3.092 10.901 13.082 make_images_data 5054 15.6 0.064 0.069 10.718 12.939 init_scf_run 11 5.9 0.000 0.001 12.146 12.147 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.146 12.147 fft_wrap_pw1pw2_140 527 13.2 3.246 3.302 11.659 11.672 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.872 9.912 dbcsr_mm_accdrv_process 20926 16.1 4.284 6.115 7.585 9.442 cp_fm_cholesky_decompose 22 10.9 9.126 9.207 9.126 9.207 wfi_extrapolate 11 7.9 0.001 0.001 9.037 9.037 grid_integrate_task_list 129 12.3 8.617 8.813 8.617 8.813 density_rs2pw 129 9.7 0.005 0.005 8.491 8.540 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.154 8.155 fft3d_ps 1301 14.7 4.002 4.009 7.538 7.586 mp_alltoall_d11v 2423 14.1 6.962 7.113 6.962 7.113 calculate_dm_sparse 129 9.5 0.001 0.001 6.772 6.809 grid_collocate_task_list 129 9.7 6.465 6.549 6.465 6.549 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.312 6.348 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.206 6.276 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=293.727000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2866.636364, yerr=147.234058 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 1.260843E+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 0.014 0.032 84.932 84.934 qs_energies 1 2.0 0.000 0.000 84.472 84.486 ls_scf 1 3.0 0.000 0.000 83.576 83.590 dbcsr_multiply_generic 111 6.7 0.014 0.016 72.578 72.753 multiply_cannon 111 7.7 0.017 0.020 55.883 57.098 multiply_cannon_loop 111 8.7 0.229 0.247 52.492 53.876 ls_scf_main 1 4.0 0.000 0.000 52.255 52.266 density_matrix_trs4 2 5.0 0.002 0.003 46.717 46.790 ls_scf_init_scf 1 4.0 0.000 0.000 28.281 28.282 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.158 27.210 mp_waitall_1 11031 10.9 22.481 25.314 22.481 25.314 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 24.998 25.013 multiply_cannon_multrec 2664 9.7 8.155 8.840 15.550 17.306 multiply_cannon_sync_h2d 2664 9.7 13.496 15.756 13.496 15.756 make_m2s 222 7.7 0.009 0.012 12.997 13.454 make_images 222 8.7 0.098 0.108 12.974 13.432 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.831 12.935 make_images_data 222 9.7 0.004 0.005 7.592 8.120 dbcsr_mm_accdrv_process 4760 10.4 0.592 0.722 7.012 8.044 hybrid_alltoall_any 227 10.6 0.215 1.838 6.571 7.830 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.393 7.347 dbcsr_mm_accdrv_process_sort 4760 11.4 6.217 7.136 6.217 7.136 calculate_norms 4752 9.8 5.504 6.218 5.504 6.218 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.048 5.180 mp_sum_l 887 5.1 3.160 4.902 3.160 4.902 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.060 3.842 mp_irecv_dv 6231 10.9 2.044 3.808 2.044 3.808 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.390 3.719 make_images_sizes 222 9.7 0.000 0.000 0.651 3.665 mp_alltoall_i44 222 10.7 0.651 3.664 0.651 3.664 arnoldi_extremal 4 6.8 0.000 0.000 3.203 3.224 arnoldi_normal_ev 4 7.8 0.001 0.003 3.203 3.224 build_subspace 16 8.4 0.009 0.012 3.108 3.110 ls_scf_post 1 4.0 0.000 0.000 3.040 3.054 ls_scf_store_result 1 5.0 0.000 0.000 2.868 2.902 dbcsr_special_finalize 555 9.7 0.005 0.006 2.400 2.727 dbcsr_merge_single_wm 555 10.7 0.466 0.621 2.392 2.718 make_images_pack 222 9.7 2.207 2.621 2.209 2.623 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.311 2.558 dbcsr_sort_data 658 11.4 2.184 2.479 2.184 2.479 dbcsr_matrix_vector_mult_local 304 10.0 2.062 2.456 2.064 2.458 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.302 2.376 buffer_matrices_ensure_size 222 8.7 1.756 2.164 1.756 2.164 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.766 1.767 rebuild_ks_matrix 3 7.3 0.000 0.000 1.757 1.758 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.757 1.758 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.934000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 2.173813E+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 0.035 0.061 90.162 90.164 qs_energies 1 2.0 0.000 0.001 89.635 89.639 ls_scf 1 3.0 0.001 0.004 88.303 88.307 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.507 74.766 multiply_cannon 111 7.7 0.027 0.035 52.508 57.558 ls_scf_main 1 4.0 0.000 0.002 54.475 54.483 multiply_cannon_loop 111 8.7 0.136 0.147 49.713 53.308 density_matrix_trs4 2 5.0 0.003 0.007 48.838 49.032 ls_scf_init_scf 1 4.0 0.000 0.002 30.081 30.085 mp_waitall_1 9105 10.9 20.574 29.862 20.574 29.862 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.883 28.974 multiply_cannon_multrec 1332 9.7 13.345 16.504 22.639 26.765 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.535 26.548 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.300 20.597 make_m2s 222 7.7 0.006 0.007 15.159 15.689 make_images 222 8.7 1.369 1.701 15.129 15.660 dbcsr_mm_accdrv_process 4041 10.4 0.369 0.557 8.893 10.562 dbcsr_mm_accdrv_process_sort 4041 11.4 8.408 10.005 8.408 10.005 make_images_data 222 9.7 0.004 0.005 8.793 9.688 hybrid_alltoall_any 227 10.6 0.541 2.555 8.147 9.470 mp_sum_l 887 5.1 5.384 8.597 5.384 8.597 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.215 7.754 mp_irecv_dv 3311 11.0 3.196 7.697 3.196 7.697 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.114 6.910 calculate_norms 2376 9.8 6.072 6.636 6.072 6.636 multiply_cannon_sync_h2d 1332 9.7 4.866 6.340 4.866 6.340 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.893 5.114 arnoldi_extremal 4 6.8 0.000 0.000 4.600 4.615 arnoldi_normal_ev 4 7.8 0.001 0.005 4.600 4.615 build_subspace 16 8.4 0.014 0.021 4.351 4.355 ls_scf_post 1 4.0 0.000 0.002 3.746 3.750 ls_scf_store_result 1 5.0 0.000 0.000 3.436 3.570 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.149 3.360 dbcsr_matrix_vector_mult_local 304 10.0 2.735 3.212 2.737 3.214 mp_allgather_i34 111 8.7 0.783 3.018 0.783 3.018 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.590 2.698 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.128 2.519 dbcsr_data_new 4174 10.1 2.114 2.387 2.114 2.387 make_images_pack 222 9.7 1.819 2.132 1.821 2.134 dbcsr_sort_data 436 11.2 1.835 2.121 1.835 2.121 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.850 1.854 rebuild_ks_matrix 3 7.3 0.000 0.000 1.838 1.841 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.010 1.838 1.841 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.164000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1780.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 2.924732E+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 0.054 0.076 92.922 92.926 qs_energies 1 2.0 0.000 0.001 92.157 92.160 ls_scf 1 3.0 0.000 0.002 90.733 90.737 dbcsr_multiply_generic 111 6.7 0.016 0.020 75.435 75.718 ls_scf_main 1 4.0 0.000 0.002 56.557 56.562 multiply_cannon 111 7.7 0.032 0.063 52.232 56.400 multiply_cannon_loop 111 8.7 0.117 0.133 49.466 53.604 density_matrix_trs4 2 5.0 0.002 0.004 50.686 50.924 mp_waitall_1 7281 11.0 23.713 33.197 23.713 33.197 ls_scf_init_scf 1 4.0 0.000 0.001 30.505 30.510 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.257 29.374 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.887 26.897 multiply_cannon_multrec 888 9.7 12.694 15.374 21.270 24.552 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.001 23.026 make_m2s 222 7.7 0.006 0.007 16.432 17.194 make_images 222 8.7 1.587 1.862 16.394 17.154 make_images_data 222 9.7 0.004 0.004 9.533 10.730 hybrid_alltoall_any 227 10.6 0.639 2.956 9.107 10.339 dbcsr_mm_accdrv_process 3754 10.4 0.317 0.492 8.094 9.353 dbcsr_mm_accdrv_process_sort 3754 11.4 7.660 8.861 7.660 8.861 mp_sum_l 887 5.1 5.095 8.757 5.095 8.757 multiply_cannon_sync_h2d 888 9.7 6.056 7.148 6.056 7.148 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.823 7.088 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.942 7.041 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.484 6.991 mp_irecv_dv 2335 11.1 2.468 6.927 2.468 6.927 arnoldi_extremal 4 6.8 0.000 0.000 5.057 5.071 arnoldi_normal_ev 4 7.8 0.002 0.010 5.057 5.071 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.733 4.985 calculate_norms 1584 9.8 4.370 4.861 4.370 4.861 build_subspace 16 8.4 0.014 0.020 4.756 4.765 mp_allgather_i34 111 8.7 0.883 3.788 0.883 3.788 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.416 3.743 ls_scf_post 1 4.0 0.000 0.002 3.670 3.674 dbcsr_matrix_vector_mult_local 304 10.0 3.016 3.581 3.018 3.583 ls_scf_store_result 1 5.0 0.000 0.000 3.423 3.491 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.852 2.957 dbcsr_data_new 4116 9.9 2.114 2.460 2.114 2.460 make_images_sizes 222 9.7 0.000 0.000 1.071 2.334 mp_alltoall_i44 222 10.7 1.071 2.334 1.071 2.334 dbcsr_sort_data 325 11.1 1.879 2.087 1.879 2.087 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.906 1.912 rebuild_ks_matrix 3 7.3 0.000 0.000 1.888 1.894 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.014 1.888 1.894 make_images_pack 222 9.7 1.624 1.886 1.627 1.889 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.926000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2286.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 3.386008E+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 0.028 0.045 96.686 96.686 qs_energies 1 2.0 0.000 0.000 96.103 96.108 ls_scf 1 3.0 0.000 0.000 94.414 94.419 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.317 78.569 ls_scf_main 1 4.0 0.000 0.000 58.682 58.683 multiply_cannon 111 7.7 0.040 0.081 51.681 55.433 density_matrix_trs4 2 5.0 0.002 0.003 52.567 52.682 multiply_cannon_loop 111 8.7 0.153 0.165 46.744 49.575 ls_scf_init_scf 1 4.0 0.000 0.000 32.532 32.534 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.330 31.402 mp_waitall_1 6369 11.0 22.793 29.984 22.793 29.984 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.863 28.873 multiply_cannon_multrec 1332 9.7 14.115 17.190 22.048 25.788 make_m2s 222 7.7 0.007 0.008 21.230 22.563 make_images 222 8.7 3.138 3.603 21.180 22.515 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.268 16.728 make_images_data 222 9.7 0.004 0.004 11.841 13.391 hybrid_alltoall_any 227 10.6 0.797 3.741 11.265 12.793 dbcsr_mm_accdrv_process 3641 10.4 0.284 0.517 7.574 9.068 dbcsr_mm_accdrv_process_sort 3641 11.4 7.124 8.572 7.124 8.572 mp_sum_l 887 5.1 3.810 7.078 3.810 7.078 multiply_cannon_sync_h2d 1332 9.7 5.503 6.153 5.503 6.153 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.102 6.006 mp_irecv_dv 3229 10.9 2.077 5.926 2.077 5.926 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.946 5.673 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.514 5.288 arnoldi_extremal 4 6.8 0.000 0.000 5.177 5.191 arnoldi_normal_ev 4 7.8 0.001 0.005 5.177 5.191 build_subspace 16 8.4 0.014 0.021 4.845 4.853 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.524 4.682 calculate_norms 2376 9.8 4.179 4.563 4.179 4.563 mp_allgather_i34 111 8.7 2.060 4.132 2.060 4.132 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 3.565 3.869 dbcsr_matrix_vector_mult_local 304 10.0 3.187 3.687 3.189 3.689 dbcsr_sort_data 658 11.4 3.103 3.428 3.103 3.428 dbcsr_special_finalize 555 9.7 0.006 0.007 2.847 3.222 dbcsr_merge_single_wm 555 10.7 0.540 0.669 2.838 3.213 ls_scf_post 1 4.0 0.000 0.000 3.201 3.207 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.019 3.060 ls_scf_store_result 1 5.0 0.000 0.000 2.958 3.011 dbcsr_data_release 10477 10.7 1.585 2.416 1.585 2.416 dbcsr_finalize 304 7.8 0.049 0.061 1.806 1.964 buffer_matrices_ensure_size 222 8.7 1.595 1.964 1.595 1.964 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.686000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2688.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 4.777320E+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 0.035 0.050 91.758 91.759 qs_energies 1 2.0 0.000 0.000 91.032 91.037 ls_scf 1 3.0 0.000 0.000 89.027 89.031 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.694 70.940 ls_scf_main 1 4.0 0.000 0.000 56.461 56.461 multiply_cannon 111 7.7 0.070 0.137 52.823 55.829 multiply_cannon_loop 111 8.7 0.088 0.097 50.229 51.738 density_matrix_trs4 2 5.0 0.002 0.003 49.555 49.603 ls_scf_init_scf 1 4.0 0.000 0.000 29.150 29.153 mp_waitall_1 5436 11.0 24.535 28.538 24.535 28.538 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.893 27.928 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.804 25.820 multiply_cannon_multrec 444 9.7 13.629 16.624 20.735 22.346 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.964 15.798 make_m2s 222 7.7 0.005 0.005 13.391 14.330 make_images 222 8.7 2.039 2.477 13.324 14.262 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.220 13.400 make_images_data 222 9.7 0.003 0.004 8.169 9.679 hybrid_alltoall_any 227 10.6 0.802 3.810 8.080 9.586 dbcsr_mm_accdrv_process 3003 10.4 0.366 0.415 6.791 7.994 multiply_cannon_sync_h2d 444 9.7 6.793 7.992 6.793 7.992 dbcsr_mm_accdrv_process_sort 3003 11.4 6.424 7.579 6.424 7.579 arnoldi_extremal 4 6.8 0.000 0.000 5.767 5.784 arnoldi_normal_ev 4 7.8 0.002 0.005 5.767 5.784 build_subspace 16 8.4 0.015 0.020 5.377 5.390 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.384 4.608 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.154 4.358 dbcsr_matrix_vector_mult_local 304 10.0 3.698 4.169 3.700 4.171 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.542 4.068 mp_irecv_dv 1241 11.2 1.530 4.036 1.530 4.036 mp_sum_l 887 5.1 2.473 4.033 2.473 4.033 mp_allgather_i34 111 8.7 1.159 3.709 1.159 3.709 calculate_norms 792 9.8 3.610 3.703 3.610 3.703 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.515 3.591 ls_scf_post 1 4.0 0.000 0.000 3.417 3.422 ls_scf_store_result 1 5.0 0.000 0.000 3.203 3.244 make_images_sizes 222 9.7 0.000 0.000 0.848 3.103 mp_alltoall_i44 222 10.7 0.847 3.102 0.847 3.102 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.702 3.010 dbcsr_data_new 4608 9.7 1.813 2.318 1.813 2.318 dbcsr_finalize 304 7.8 0.062 0.077 2.194 2.290 dbcsr_merge_all 275 8.9 0.479 0.532 2.053 2.135 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.035 2.036 rebuild_ks_matrix 3 7.3 0.000 0.000 2.003 2.004 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.003 2.004 qs_energies_init_hamiltonians 1 3.0 0.003 0.022 1.989 1.989 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.759000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3752.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 8.894636E+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 0.098 0.127 107.311 107.312 qs_energies 1 2.0 0.000 0.000 105.894 105.903 ls_scf 1 3.0 0.000 0.000 102.924 102.933 dbcsr_multiply_generic 111 6.7 0.023 0.027 76.520 76.672 ls_scf_main 1 4.0 0.000 0.000 63.738 63.739 density_matrix_trs4 2 5.0 0.002 0.003 54.703 54.762 multiply_cannon 111 7.7 0.162 0.349 50.276 52.672 multiply_cannon_loop 111 8.7 0.100 0.103 47.165 47.764 ls_scf_init_scf 1 4.0 0.000 0.000 35.419 35.420 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.860 33.887 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 31.140 31.147 mp_waitall_1 4527 11.1 21.966 25.656 21.966 25.656 make_m2s 222 7.7 0.005 0.005 22.583 23.794 make_images 222 8.7 3.570 3.884 22.475 23.686 multiply_cannon_multrec 444 9.7 17.855 18.622 22.492 23.280 hybrid_alltoall_any 227 10.6 1.657 3.631 12.803 15.706 make_images_data 222 9.7 0.004 0.004 13.012 15.257 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.383 12.117 multiply_cannon_sync_h2d 444 9.7 8.789 8.845 8.789 8.845 arnoldi_extremal 4 6.8 0.000 0.000 7.391 7.402 arnoldi_normal_ev 4 7.8 0.003 0.009 7.391 7.402 build_subspace 16 8.4 0.026 0.036 6.841 6.854 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.456 5.635 dbcsr_matrix_vector_mult_local 304 10.0 5.046 5.378 5.048 5.381 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.168 5.265 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.871 5.131 dbcsr_mm_accdrv_process 1814 10.4 0.314 0.357 4.462 4.594 dbcsr_mm_accdrv_process_sort 1814 11.4 4.124 4.259 4.124 4.259 mp_allgather_i34 111 8.7 1.230 3.946 1.230 3.946 ls_scf_post 1 4.0 0.000 0.000 3.767 3.777 make_images_sizes 222 9.7 0.000 0.000 1.448 3.574 mp_alltoall_i44 222 10.7 1.447 3.573 1.447 3.573 ls_scf_store_result 1 5.0 0.000 0.000 3.490 3.524 calculate_norms 792 9.8 3.232 3.268 3.232 3.268 dbcsr_finalize 304 7.8 0.082 0.089 3.086 3.194 dbcsr_merge_all 275 8.9 0.883 0.915 2.864 2.970 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.939 2.939 dbcsr_complete_redistribute 5 7.6 1.432 1.471 2.743 2.857 dbcsr_sort_data 325 11.1 2.437 2.520 2.437 2.520 matrix_ls_to_qs 2 6.0 0.000 0.000 2.398 2.513 compute_matrix_preconditioner 1 6.0 0.002 0.002 2.498 2.506 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.455 2.456 rebuild_ks_matrix 3 7.3 0.000 0.000 2.389 2.391 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.389 2.391 dbcsr_new_transposed 4 7.5 0.243 0.255 2.283 2.302 dbcsr_data_new 6591 9.6 1.830 2.285 1.830 2.285 dbcsr_frobenius_norm 74 6.6 2.057 2.136 2.215 2.253 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.208 dbcsr_add_anytype 103 7.2 0.859 0.892 2.131 2.207 dbcsr_data_release 12724 10.6 1.985 2.171 1.985 2.171 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.312000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7029.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1df1199c145652d4a172ffedffd9b9e59ce98d27_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 591.679488E+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 993462357. 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 0.022 0.054 220.246 220.248 qs_mol_dyn_low 1 2.0 0.003 0.008 219.234 219.256 qs_forces 5 3.8 0.005 0.035 219.146 219.148 qs_energies 5 4.8 0.002 0.009 216.217 216.243 scf_env_do_scf 5 5.8 0.000 0.006 203.225 203.227 scf_env_do_scf_inner_loop 105 6.6 0.002 0.007 177.020 177.022 qs_scf_new_mos 105 7.6 0.000 0.001 138.374 138.506 qs_scf_loop_do_ot 105 8.6 0.001 0.001 138.373 138.506 dbcsr_multiply_generic 1445 12.2 0.129 0.140 130.577 130.985 ot_scf_mini 105 9.6 0.003 0.004 128.569 128.719 multiply_cannon 1445 13.2 0.275 0.285 112.179 113.697 multiply_cannon_loop 1445 14.2 2.873 3.031 110.579 111.793 velocity_verlet 4 3.0 0.001 0.002 104.760 104.761 ot_mini 105 10.6 0.001 0.003 59.514 59.640 multiply_cannon_multrec 69360 15.2 29.598 34.454 39.533 44.514 mp_waitall_1 488190 16.1 33.582 40.056 33.582 40.056 qs_ot_get_p 112 10.4 0.001 0.002 39.283 39.614 qs_ot_get_derivative 55 11.6 0.001 0.002 37.864 37.964 multiply_cannon_sync_h2d 69360 15.2 29.121 32.775 29.121 32.775 multiply_cannon_metrocomm3 69360 15.2 0.199 0.214 25.184 31.205 qs_ot_p2m_diag 40 11.0 0.020 0.034 28.667 28.783 rebuild_ks_matrix 110 8.4 0.000 0.000 28.236 28.417 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.023 28.235 28.416 init_scf_loop 7 6.6 0.001 0.016 26.176 26.178 qs_ks_update_qs_env 112 7.6 0.001 0.001 25.902 26.064 cp_dbcsr_syevd 40 12.0 0.002 0.003 25.494 25.495 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 22.911 23.144 apply_single 62 13.6 0.000 0.000 22.911 23.144 prepare_preconditioner 7 7.6 0.000 0.000 21.396 21.433 make_preconditioner 7 8.6 0.000 0.004 21.396 21.433 ot_new_cg_direction 55 11.6 0.001 0.002 20.946 20.946 cp_fm_syevd 40 13.0 0.000 0.001 20.492 20.643 qs_rho_update_rho_low 110 7.6 0.000 0.004 16.092 16.455 calculate_rho_elec 110 8.6 0.030 0.036 16.092 16.454 cp_fm_redistribute_end 40 14.0 7.938 15.838 7.942 15.839 cp_fm_syevd_base 40 14.0 7.890 15.792 7.890 15.792 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.619 14.790 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.413 14.485 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.789 13.873 mp_sum_l 4764 12.2 11.835 12.664 11.835 12.664 calculate_dm_sparse 110 9.5 0.000 0.001 11.335 11.529 pw_transfer 1645 12.4 0.079 0.102 11.226 11.399 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.088 11.265 density_rs2pw 110 9.6 0.005 0.006 10.417 10.862 dbcsr_mm_accdrv_process 154766 15.8 6.332 6.512 9.803 10.629 init_scf_run 5 5.8 0.000 0.001 10.362 10.363 scf_env_initial_rho_setup 5 6.8 0.001 0.006 10.362 10.363 cp_fm_cholesky_invert 7 10.6 10.182 10.189 10.182 10.189 qs_vxc_create 110 10.4 0.003 0.027 10.145 10.183 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.004 10.086 fft_wrap_pw1pw2_240 915 15.0 0.841 0.914 9.721 9.911 check_diag 80 13.5 8.594 8.886 9.331 9.482 sum_up_and_integrate 60 10.3 0.001 0.004 8.149 8.159 integrate_v_rspace 60 11.3 0.001 0.002 8.132 8.143 fft3d_pb 915 16.0 2.398 2.575 7.891 8.069 transfer_rs2pw 445 10.6 0.007 0.009 7.288 7.673 xc_rho_set_and_dset_create 110 12.4 0.076 0.095 7.192 7.430 multiply_cannon_metrocomm1 69360 15.2 0.097 0.103 4.339 6.828 xc_vxc_pw_create 60 11.3 0.039 0.052 6.772 6.810 make_full_single_inverse 7 9.6 0.001 0.007 6.726 6.764 acc_transpose_blocks 69360 15.2 0.361 0.377 6.434 6.705 make_m2s 2890 13.2 0.079 0.091 6.066 6.627 make_images 2890 14.2 0.238 0.259 5.957 6.518 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 6.036 6.054 xc_pw_derive 510 13.4 0.006 0.007 5.965 6.026 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 5.620 5.667 calculate_first_density_matrix 1 7.0 0.001 0.007 5.563 5.587 mp_alltoall_z22v 2340 17.7 4.874 5.153 4.874 5.153 multiply_cannon_metrocomm4 67915 15.2 0.185 0.201 2.019 5.079 mp_waitany 7680 13.5 4.333 4.845 4.333 4.845 mp_irecv_dv 143422 16.1 1.710 4.503 1.710 4.503 wfi_extrapolate 5 7.8 0.000 0.002 4.480 4.480 potential_pw2rs 60 12.3 0.002 0.003 4.415 4.442 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=220.248000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.200000, yerr=2.785678 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 1df1199c145652d4a172ffedffd9b9e59ce98d27 Summary: empty Status: OK