=== 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: d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb ################# 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 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (31.05.2023) # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/01 job id: 47035530 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/02 job id: 47035532 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/03 job id: 47035535 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/04 job id: 47035537 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/05 job id: 47035540 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/06 job id: 47035547 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/07 job id: 47035553 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/08 job id: 47035555 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/09 job id: 47035557 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/10 job id: 47035559 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/11 job id: 47035561 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/12 job id: 47035563 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/13 job id: 47035566 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/14 job id: 47035568 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/15 job id: 47035572 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/16 job id: 47035574 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/17 job id: 47035576 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/18 job id: 47035578 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/19 job id: 47035580 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/20 job id: 47035583 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/21 job id: 47035585 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/22 job id: 47035588 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/23 job id: 47035590 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/24 job id: 47035593 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/25 job id: 47035597 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/26 job id: 47035601 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/27 job id: 47035607 --- 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/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 0.023 0.034 135.341 135.342 farming_run 1 2.0 134.656 134.657 135.308 135.311 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.463312E+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 1103589. 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 0.024 0.044 115.727 115.727 qs_energies 1 2.0 0.000 0.000 115.044 115.045 mp2_main 1 3.0 0.000 0.000 112.957 112.958 mp2_gpw_main 1 4.0 0.021 0.035 111.932 111.933 mp2_ri_gpw_compute_in 1 5.0 0.172 0.182 92.935 93.290 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.955 55.324 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.397 46.547 get_2c_integrals 1 6.0 0.008 0.009 37.206 37.789 integrate_v_rspace 273 8.0 0.435 0.449 24.778 29.810 pw_transfer 6555 10.6 0.373 0.397 27.366 28.009 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 26.043 26.505 grid_integrate_task_list 273 9.0 20.587 26.114 20.587 26.114 fft_wrap_pw1pw2_100 2178 12.4 1.195 1.406 23.559 24.038 compute_2c_integrals 1 7.0 0.002 0.002 19.645 19.649 compute_2c_integrals_loop_lm 1 8.0 0.003 0.008 18.842 19.376 mp2_eri_2c_integrate_gpw 1 9.0 2.392 2.463 18.838 19.374 rpa_ri_compute_en 1 5.0 0.019 0.021 18.887 19.061 cp_fm_cholesky_decompose 12 8.2 17.531 18.155 17.531 18.155 cholesky_decomp 1 7.0 0.000 0.000 16.403 17.037 fft3d_s 5443 13.4 16.136 16.399 16.158 16.421 ao_to_mo_and_store_B_mult_1 272 7.0 10.722 15.297 10.722 15.297 calculate_wavefunction 272 8.0 5.407 5.554 12.539 13.152 rpa_num_int 1 6.0 0.000 0.001 10.694 10.703 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.652 10.684 calc_mat_Q 8 8.0 0.000 0.000 9.481 9.578 contract_S_to_Q 8 9.0 0.000 0.000 8.901 8.998 calc_potential_gpw 544 9.5 0.005 0.006 8.233 8.627 parallel_gemm_fm 14 9.1 0.000 0.000 8.504 8.604 parallel_gemm_fm_cosma 14 10.1 8.504 8.604 8.504 8.604 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.186 8.507 potential_pw2rs 545 10.0 0.108 0.112 7.687 8.342 collocate_single_gaussian 272 10.0 0.040 0.042 7.453 7.777 create_integ_mat 1 6.0 0.014 0.026 7.730 7.739 array2fm 1 7.0 0.000 0.000 6.736 7.131 pw_scatter_s 2720 13.7 4.442 4.645 4.442 4.645 pw_gather_s 2722 13.2 3.863 4.207 3.863 4.207 array2fm_buffer_send 1 8.0 3.008 3.173 3.008 3.173 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.932078, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2796.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.040 400.384 400.385 farming_run 1 2.0 399.558 399.569 400.343 400.347 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 897827128576 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 249788821 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 1.227256E+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 743 386399. 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 0.010 0.031 211.405 211.405 qs_energies 1 2.0 0.003 0.012 211.173 211.199 scf_env_do_scf 1 3.0 0.000 0.000 107.460 107.460 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.552 106.560 rebuild_ks_matrix 4 6.0 0.000 0.000 106.551 106.559 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.064 106.551 106.559 hfx_ks_matrix 4 8.0 0.001 0.001 106.164 106.168 integrate_four_center 4 9.0 0.144 0.453 106.163 106.167 mp2_main 1 3.0 0.007 0.035 103.405 103.423 mp2_gpw_main 1 4.0 0.046 0.106 102.488 102.500 integrate_four_center_main 4 10.0 0.122 0.506 96.684 100.082 integrate_four_center_bin 267 11.0 96.563 99.949 96.563 99.949 init_scf_loop 1 4.0 0.000 0.000 93.295 93.295 mp2_ri_gpw_compute_in 1 5.0 0.175 0.512 74.769 75.894 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.029 54.060 55.178 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.163 41.634 46.771 integrate_v_rspace 95 8.0 0.398 0.577 28.023 32.991 pw_transfer 2240 10.6 0.143 0.159 29.925 30.299 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.938 29.285 mp2_ri_gpw_compute_en 1 5.0 0.063 0.108 27.468 29.201 grid_integrate_task_list 95 9.0 23.306 28.482 23.306 28.482 ao_to_mo_and_store_B_mult_1 91 7.0 10.738 27.320 10.738 27.320 fft_wrap_pw1pw2_100 730 12.4 1.274 1.452 26.659 27.063 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.855 2.289 25.438 25.450 get_2c_integrals 1 6.0 0.008 0.053 20.503 20.650 compute_2c_integrals 1 7.0 0.003 0.004 19.479 19.604 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 18.878 19.218 mp2_eri_2c_integrate_gpw 1 9.0 1.741 1.850 18.877 19.216 fft3d_s 1823 13.4 18.446 18.894 18.459 18.907 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.163 14.163 calculate_wavefunction 91 8.0 2.011 2.047 9.731 9.976 potential_pw2rs 186 10.0 0.034 0.035 8.644 9.191 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.589 8.701 9.023 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.514 8.472 8.775 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.239 8.589 local_gemm 172 8.0 8.145 8.460 8.145 8.460 collocate_single_gaussian 91 10.0 0.031 0.244 7.903 8.170 calc_potential_gpw 182 9.5 0.002 0.002 7.906 8.113 mp_sync 37 10.5 4.100 6.858 4.100 6.858 mp_sendrecv_dm3 2068 8.0 6.509 6.822 6.509 6.822 mp2_ri_gpw_compute_en_ener 172 7.0 6.366 6.427 6.366 6.427 pw_gather_s 912 13.2 4.907 5.363 4.907 5.363 pw_scatter_s 910 13.7 3.920 4.311 3.920 4.311 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.491451, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 451.338240E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.026 51.993 51.994 qs_mol_dyn_low 1 2.0 0.003 0.003 51.784 51.791 qs_forces 11 3.9 0.002 0.003 51.721 51.722 qs_energies 11 4.9 0.006 0.008 50.223 50.236 scf_env_do_scf 11 5.9 0.000 0.001 43.586 43.587 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 41.456 41.457 dbcsr_multiply_generic 2286 12.5 0.094 0.098 31.158 31.738 qs_scf_new_mos 108 7.5 0.000 0.001 31.147 31.422 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.147 31.422 ot_scf_mini 108 9.5 0.002 0.002 29.559 29.745 multiply_cannon 2286 13.5 0.182 0.191 24.962 26.660 multiply_cannon_loop 2286 14.5 1.450 1.532 24.255 25.959 velocity_verlet 10 3.0 0.001 0.002 25.258 25.259 ot_mini 108 10.5 0.001 0.001 17.355 17.597 qs_ot_get_derivative 108 11.5 0.001 0.001 14.486 14.677 mp_waitall_1 245248 16.5 7.378 13.559 7.378 13.559 multiply_cannon_metrocomm3 54864 15.5 0.069 0.075 5.249 12.157 multiply_cannon_multrec 54864 15.5 4.378 6.886 7.665 11.119 rebuild_ks_matrix 119 8.3 0.000 0.000 8.015 8.153 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.014 8.152 qs_ot_get_p 119 10.4 0.001 0.001 7.797 8.070 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.050 7.175 multiply_cannon_sync_h2d 54864 15.5 6.048 6.626 6.048 6.626 mp_sum_l 7287 12.8 4.092 5.832 4.092 5.832 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.010 5.438 init_scf_run 11 5.9 0.000 0.001 5.350 5.350 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.350 5.350 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.125 5.169 sum_up_and_integrate 119 10.3 0.012 0.014 4.625 4.632 integrate_v_rspace 119 11.3 0.002 0.002 4.613 4.620 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.475 4.570 dbcsr_mm_accdrv_process 76910 16.1 1.078 1.782 3.208 4.510 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.444 4.444 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.348 4.430 calculate_rho_elec 119 8.7 0.012 0.017 4.348 4.429 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.238 4.239 cp_fm_redistribute_end 50 14.0 2.163 4.211 2.174 4.218 cp_fm_diag_elpa_base 50 14.0 2.038 4.097 2.042 4.106 calculate_first_density_matrix 1 7.0 0.000 0.000 3.153 3.165 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.920 3.118 apply_single 119 13.6 0.000 0.000 2.920 3.118 calculate_dm_sparse 119 9.5 0.000 0.001 2.938 3.048 rs_pw_transfer 974 11.9 0.011 0.013 2.833 2.957 jit_kernel_multiply 13 15.8 2.073 2.653 2.073 2.653 ot_diis_step 108 11.5 0.006 0.006 2.633 2.634 multiply_cannon_metrocomm1 54864 15.5 0.054 0.058 1.393 2.515 density_rs2pw 119 9.7 0.004 0.005 2.324 2.438 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.232 2.271 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.209 2.212 acc_transpose_blocks 54864 15.5 0.216 0.241 1.692 2.201 grid_integrate_task_list 119 12.3 2.019 2.113 2.019 2.113 init_scf_loop 11 6.9 0.000 0.000 2.111 2.111 wfi_extrapolate 11 7.9 0.001 0.001 2.106 2.106 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.928 1.981 potential_pw2rs 119 12.3 0.004 0.004 1.950 1.960 pw_transfer 1439 11.6 0.052 0.056 1.857 1.931 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.782 1.858 make_m2s 4572 13.5 0.053 0.056 1.788 1.841 mp_sum_d 4135 12.0 1.190 1.822 1.190 1.822 make_images 4572 14.5 0.133 0.138 1.706 1.759 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.506 1.542 parallel_gemm_fm 81 9.0 0.000 0.000 1.532 1.536 parallel_gemm_fm_cosma 81 10.0 1.532 1.536 1.532 1.536 fft3d_ps 1201 14.6 0.375 0.481 1.423 1.499 mp_waitany 12084 13.8 1.294 1.487 1.294 1.487 mp_alltoall_d11v 2130 13.8 1.340 1.465 1.340 1.465 fft_wrap_pw1pw2_140 487 13.2 0.185 0.207 1.367 1.443 grid_collocate_task_list 119 9.7 1.349 1.426 1.349 1.426 make_basis_sm 11 9.8 0.000 0.000 1.085 1.087 dbcsr_dot_sd 1205 11.9 0.049 0.061 0.638 1.048 acc_transpose_blocks_kernels 54864 16.5 0.232 0.370 0.757 1.045 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=51.994000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.454545, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 486.219776E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1030290. 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 0.037 0.062 39.902 39.905 qs_mol_dyn_low 1 2.0 0.009 0.028 39.400 39.407 qs_forces 11 3.9 0.003 0.004 38.957 38.961 qs_energies 11 4.9 0.003 0.014 37.252 37.258 scf_env_do_scf 11 5.9 0.001 0.003 32.027 32.028 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.507 29.508 dbcsr_multiply_generic 2286 12.5 0.101 0.103 21.447 21.814 qs_scf_new_mos 108 7.5 0.001 0.001 20.278 20.512 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.278 20.511 ot_scf_mini 108 9.5 0.003 0.003 19.364 19.533 velocity_verlet 10 3.0 0.001 0.002 18.745 18.759 multiply_cannon 2286 13.5 0.207 0.214 16.358 17.991 multiply_cannon_loop 2286 14.5 0.905 0.972 15.239 16.797 ot_mini 108 10.5 0.001 0.001 11.860 12.096 mp_waitall_1 200699 16.5 5.677 10.851 5.677 10.851 qs_ot_get_derivative 108 11.5 0.001 0.001 9.384 9.554 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 4.154 9.331 multiply_cannon_multrec 27432 15.5 1.992 4.287 5.875 8.576 rebuild_ks_matrix 119 8.3 0.000 0.000 7.255 7.399 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.255 7.398 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.400 6.532 dbcsr_mm_accdrv_process 47894 16.0 2.934 4.781 3.813 5.644 qs_ot_get_p 119 10.4 0.001 0.001 4.590 4.824 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.617 4.487 sum_up_and_integrate 119 10.3 0.024 0.027 4.217 4.223 integrate_v_rspace 119 11.3 0.002 0.002 4.193 4.201 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.081 4.155 apply_single 119 13.6 0.000 0.000 3.081 4.155 mp_sum_l 7287 12.8 2.153 4.078 2.153 4.078 init_scf_run 11 5.9 0.000 0.001 3.959 3.960 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.959 3.960 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.848 3.882 calculate_rho_elec 119 8.7 0.021 0.024 3.848 3.882 qs_ot_p2m_diag 50 11.0 0.013 0.044 3.123 3.141 rs_pw_transfer 974 11.9 0.010 0.011 2.606 3.079 multiply_cannon_sync_h2d 27432 15.5 2.176 2.788 2.176 2.788 make_m2s 4572 13.5 0.052 0.054 2.505 2.722 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.674 2.679 make_images 4572 14.5 0.200 0.236 2.415 2.631 density_rs2pw 119 9.7 0.004 0.004 2.123 2.604 init_scf_loop 11 6.9 0.001 0.004 2.492 2.492 ot_diis_step 108 11.5 0.011 0.011 2.422 2.423 calculate_first_density_matrix 1 7.0 0.000 0.002 2.418 2.421 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.317 2.318 cp_fm_redistribute_end 50 14.0 1.176 2.291 1.180 2.295 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.152 2.239 cp_fm_diag_elpa_base 50 14.0 1.082 2.191 1.110 2.232 calculate_dm_sparse 119 9.5 0.000 0.001 2.050 2.126 pw_transfer 1439 11.6 0.065 0.070 1.968 2.000 potential_pw2rs 119 12.3 0.006 0.006 1.935 1.948 grid_integrate_task_list 119 12.3 1.840 1.944 1.840 1.944 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.937 1.940 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.876 1.910 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.726 1.771 jit_kernel_multiply 9 16.1 0.827 1.733 0.827 1.733 make_images_data 4572 15.5 0.044 0.050 1.158 1.581 prepare_preconditioner 11 7.9 0.000 0.000 1.545 1.572 make_preconditioner 11 8.9 0.000 0.001 1.545 1.572 acc_transpose_blocks 27432 15.5 0.109 0.115 1.222 1.518 fft_wrap_pw1pw2_140 487 13.2 0.204 0.213 1.466 1.500 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.436 1.496 wfi_extrapolate 11 7.9 0.001 0.001 1.474 1.474 fft3d_ps 1201 14.6 0.525 0.582 1.424 1.453 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.013 1.452 grid_collocate_task_list 119 9.7 1.273 1.421 1.273 1.421 mp_allgather_i34 2286 14.5 0.560 1.326 0.560 1.326 mp_alltoall_d11v 2130 13.8 1.203 1.319 1.203 1.319 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.248 1.255 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.192 1.237 mp_sum_d 4135 12.0 0.654 1.117 0.654 1.117 mp_waitany 5720 13.7 0.554 1.062 0.554 1.062 rs_pw_transfer_RS2PW_140 130 11.5 0.137 0.146 0.576 1.049 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.975 0.979 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.915 0.928 acc_transpose_blocks_kernels 27432 16.5 0.183 0.271 0.670 0.888 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.905000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.090909, yerr=1.504813 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 520.667136E+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 63497. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 830269. 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 0.011 0.026 33.835 33.836 qs_mol_dyn_low 1 2.0 0.003 0.003 33.585 33.592 qs_forces 11 3.9 0.002 0.003 33.514 33.515 qs_energies 11 4.9 0.001 0.002 31.907 31.909 scf_env_do_scf 11 5.9 0.001 0.001 26.306 26.306 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.710 23.711 dbcsr_multiply_generic 2286 12.5 0.097 0.099 17.322 17.413 velocity_verlet 10 3.0 0.003 0.019 15.621 15.622 qs_scf_new_mos 108 7.5 0.001 0.001 15.374 15.389 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.373 15.388 multiply_cannon 2286 13.5 0.195 0.208 13.943 14.716 ot_scf_mini 108 9.5 0.002 0.002 14.635 14.645 multiply_cannon_loop 2286 14.5 0.637 0.667 13.153 13.922 ot_mini 108 10.5 0.001 0.001 9.079 9.095 qs_ot_get_derivative 108 11.5 0.001 0.001 7.560 7.573 multiply_cannon_multrec 18288 15.5 1.980 2.993 6.971 7.393 rebuild_ks_matrix 119 8.3 0.000 0.000 6.478 6.496 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.478 6.495 dbcsr_mm_accdrv_process 38222 16.0 4.080 5.776 4.906 5.943 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.708 5.723 init_scf_run 11 5.9 0.000 0.001 4.348 4.348 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.348 4.348 mp_waitall_1 158411 16.6 3.205 4.152 3.205 4.152 sum_up_and_integrate 119 10.3 0.030 0.031 4.027 4.034 integrate_v_rspace 119 11.3 0.002 0.003 3.997 4.007 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.012 3.654 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.570 3.574 calculate_rho_elec 119 8.7 0.031 0.031 3.570 3.574 qs_ot_get_p 119 10.4 0.001 0.001 3.358 3.375 calculate_first_density_matrix 1 7.0 0.000 0.000 3.070 3.071 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.833 2.817 rs_pw_transfer 974 11.9 0.009 0.010 2.409 2.674 init_scf_loop 11 6.9 0.000 0.000 2.577 2.579 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.068 2.372 apply_single 119 13.6 0.000 0.000 2.068 2.371 calculate_dm_sparse 119 9.5 0.000 0.000 2.312 2.325 density_rs2pw 119 9.7 0.004 0.004 2.052 2.307 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.277 2.282 jit_kernel_multiply 10 15.9 0.775 2.016 0.775 2.016 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.009 2.009 pw_transfer 1439 11.6 0.065 0.070 1.964 1.978 make_m2s 4572 13.5 0.044 0.045 1.826 1.970 grid_integrate_task_list 119 12.3 1.801 1.894 1.801 1.894 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.871 1.888 make_images 4572 14.5 0.190 0.203 1.740 1.881 prepare_preconditioner 11 7.9 0.000 0.000 1.781 1.784 make_preconditioner 11 8.9 0.000 0.000 1.781 1.784 potential_pw2rs 119 12.3 0.007 0.008 1.773 1.782 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.777 1.778 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.751 1.760 cp_fm_diag_elpa_base 50 14.0 1.728 1.736 1.749 1.759 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.633 1.720 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.666 1.671 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.630 1.636 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.595 1.602 acc_transpose_blocks 18288 15.5 0.076 0.078 1.368 1.582 multiply_cannon_sync_h2d 18288 15.5 1.389 1.551 1.389 1.551 fft_wrap_pw1pw2_140 487 13.2 0.255 0.260 1.488 1.504 ot_diis_step 108 11.5 0.011 0.011 1.489 1.490 mp_sum_l 7287 12.8 1.161 1.455 1.161 1.455 grid_collocate_task_list 119 9.7 1.245 1.409 1.245 1.409 fft3d_ps 1201 14.6 0.533 0.550 1.355 1.373 wfi_extrapolate 11 7.9 0.001 0.001 1.222 1.222 acc_transpose_blocks_kernels 18288 16.5 0.210 0.219 0.912 1.117 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.031 1.032 make_images_data 4572 15.5 0.044 0.048 0.821 0.981 multiply_cannon_metrocomm1 18288 15.5 0.028 0.030 0.544 0.977 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.906 0.925 jit_kernel_transpose 5 15.6 0.702 0.907 0.702 0.907 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.715 0.905 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.169 0.616 0.885 mp_waitany 9880 13.7 0.620 0.880 0.620 0.880 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.843 0.846 mp_alltoall_d11v 2130 13.8 0.710 0.809 0.710 0.809 cp_fm_cholesky_invert 11 10.9 0.750 0.754 0.750 0.754 mp_alltoall_z22v 1201 16.6 0.674 0.739 0.674 0.739 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.658 0.727 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.836000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.000000, yerr=1.651446 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 555.540480E+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 63496. 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 0.012 0.026 36.338 36.339 qs_mol_dyn_low 1 2.0 0.003 0.003 36.014 36.022 qs_forces 11 3.9 0.002 0.002 35.934 35.935 qs_energies 11 4.9 0.002 0.003 34.201 34.209 scf_env_do_scf 11 5.9 0.001 0.001 28.725 28.726 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.284 25.284 dbcsr_multiply_generic 2286 12.5 0.104 0.107 18.796 18.923 velocity_verlet 10 3.0 0.004 0.023 17.992 17.994 qs_scf_new_mos 108 7.5 0.001 0.001 16.763 16.825 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.763 16.825 ot_scf_mini 108 9.5 0.002 0.003 15.818 15.874 multiply_cannon 2286 13.5 0.225 0.261 15.180 15.619 multiply_cannon_loop 2286 14.5 0.946 0.980 14.235 14.568 ot_mini 108 10.5 0.001 0.001 9.737 9.808 multiply_cannon_multrec 27432 15.5 2.653 3.415 9.277 9.657 dbcsr_mm_accdrv_process 47916 15.9 5.763 7.617 6.534 8.032 qs_ot_get_derivative 108 11.5 0.001 0.001 7.903 7.958 rebuild_ks_matrix 119 8.3 0.000 0.000 6.606 6.664 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.606 6.663 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.857 5.907 init_scf_run 11 5.9 0.000 0.001 4.144 4.145 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.144 4.145 sum_up_and_integrate 119 10.3 0.035 0.037 3.853 3.862 integrate_v_rspace 119 11.3 0.003 0.003 3.818 3.827 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.592 3.624 calculate_rho_elec 119 8.7 0.040 0.046 3.592 3.624 qs_ot_get_p 119 10.4 0.001 0.001 3.465 3.547 init_scf_loop 11 6.9 0.000 0.000 3.419 3.419 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.853 3.311 calculate_first_density_matrix 1 7.0 0.000 0.000 2.767 2.769 prepare_preconditioner 11 7.9 0.000 0.000 2.574 2.582 make_preconditioner 11 8.9 0.000 0.000 2.574 2.582 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.101 2.549 apply_single 119 13.6 0.000 0.000 2.101 2.549 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.181 2.508 mp_waitall_1 137007 16.6 1.758 2.396 1.758 2.396 make_m2s 4572 13.5 0.054 0.055 2.175 2.290 rs_pw_transfer 974 11.9 0.009 0.009 2.029 2.209 calculate_dm_sparse 119 9.5 0.000 0.000 2.146 2.203 make_images 4572 14.5 0.271 0.333 2.068 2.182 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.172 2.181 density_rs2pw 119 9.7 0.004 0.004 1.981 2.151 jit_kernel_multiply 10 15.8 0.711 2.141 0.711 2.141 pw_transfer 1439 11.6 0.065 0.069 2.042 2.075 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.950 1.987 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.963 1.964 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.929 1.957 grid_integrate_task_list 119 12.3 1.830 1.913 1.830 1.913 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.852 1.853 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.792 1.804 ot_diis_step 108 11.5 0.012 0.012 1.789 1.790 fft_wrap_pw1pw2_140 487 13.2 0.290 0.304 1.589 1.628 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.604 1.613 cp_fm_diag_elpa_base 50 14.0 1.570 1.588 1.602 1.611 acc_transpose_blocks 27432 15.5 0.111 0.114 1.565 1.602 potential_pw2rs 119 12.3 0.008 0.009 1.596 1.600 mp_sum_l 7287 12.8 1.007 1.543 1.007 1.543 fft3d_ps 1201 14.6 0.557 0.606 1.386 1.412 grid_collocate_task_list 119 9.7 1.248 1.340 1.248 1.340 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.325 1.335 wfi_extrapolate 11 7.9 0.001 0.001 1.324 1.324 multiply_cannon_metrocomm3 27432 15.5 0.038 0.040 0.758 1.301 cp_fm_upper_to_full 72 14.2 0.814 1.161 0.814 1.161 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.108 1.125 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.067 1.068 dbcsr_complete_redistribute 329 12.2 0.121 0.150 0.766 1.039 multiply_cannon_sync_h2d 27432 15.5 0.987 1.039 0.987 1.039 make_images_data 4572 15.5 0.044 0.048 0.839 0.962 acc_transpose_blocks_kernels 27432 16.5 0.270 0.278 0.916 0.940 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.732 0.894 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.872 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.839 0.844 mp_alltoall_d11v 2130 13.8 0.742 0.844 0.742 0.844 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.567 0.832 cp_fm_cholesky_invert 11 10.9 0.745 0.748 0.745 0.748 mp_alltoall_z22v 1201 16.6 0.712 0.740 0.712 0.740 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.339000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.727273, yerr=3.193485 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 600.498176E+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 63495. 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 0.013 0.027 28.360 28.361 qs_mol_dyn_low 1 2.0 0.003 0.003 28.179 28.186 qs_forces 11 3.9 0.002 0.003 28.119 28.122 qs_energies 11 4.9 0.002 0.002 26.396 26.400 scf_env_do_scf 11 5.9 0.001 0.002 21.675 21.675 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.166 19.166 velocity_verlet 10 3.0 0.001 0.001 14.455 14.458 dbcsr_multiply_generic 2286 12.5 0.092 0.098 12.369 12.440 qs_scf_new_mos 108 7.5 0.001 0.001 11.159 11.184 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.158 11.183 ot_scf_mini 108 9.5 0.002 0.002 10.479 10.506 multiply_cannon 2286 13.5 0.235 0.243 9.848 10.314 multiply_cannon_loop 2286 14.5 0.331 0.344 8.951 9.144 multiply_cannon_multrec 9144 15.5 1.916 2.149 6.078 6.308 rebuild_ks_matrix 119 8.3 0.000 0.000 5.986 6.009 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.986 6.009 ot_mini 108 10.5 0.001 0.001 5.947 5.980 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.322 5.343 qs_ot_get_derivative 108 11.5 0.001 0.001 4.640 4.666 dbcsr_mm_accdrv_process 12550 15.8 3.263 4.091 4.061 4.132 sum_up_and_integrate 119 10.3 0.037 0.041 3.680 3.685 integrate_v_rspace 119 11.3 0.003 0.003 3.643 3.648 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.544 3.552 calculate_rho_elec 119 8.7 0.060 0.061 3.544 3.552 init_scf_run 11 5.9 0.000 0.001 3.287 3.287 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.287 3.287 qs_ot_get_p 119 10.4 0.001 0.001 2.768 2.812 init_scf_loop 11 6.9 0.000 0.000 2.489 2.490 calculate_first_density_matrix 1 7.0 0.000 0.000 2.148 2.149 pw_transfer 1439 11.6 0.066 0.068 2.068 2.078 density_rs2pw 119 9.7 0.004 0.004 1.840 1.987 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.974 1.986 grid_integrate_task_list 119 12.3 1.848 1.926 1.848 1.926 mp_waitall_1 115863 16.7 1.409 1.904 1.409 1.904 make_m2s 4572 13.5 0.034 0.035 1.734 1.889 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.834 1.837 rs_pw_transfer 974 11.9 0.008 0.008 1.672 1.798 make_images 4572 14.5 0.267 0.304 1.644 1.798 prepare_preconditioner 11 7.9 0.000 0.000 1.763 1.768 make_preconditioner 11 8.9 0.000 0.000 1.763 1.768 calculate_dm_sparse 119 9.5 0.000 0.000 1.710 1.729 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.653 1.682 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.634 1.634 fft_wrap_pw1pw2_140 487 13.2 0.366 0.375 1.617 1.630 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.575 1.576 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.515 1.532 jit_kernel_multiply 8 15.9 0.760 1.433 0.760 1.433 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.422 1.430 grid_collocate_task_list 119 9.7 1.296 1.413 1.296 1.413 potential_pw2rs 119 12.3 0.010 0.011 1.404 1.408 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.354 1.362 cp_fm_diag_elpa_base 50 14.0 1.327 1.344 1.352 1.360 fft3d_ps 1201 14.6 0.562 0.571 1.304 1.315 ot_diis_step 108 11.5 0.012 0.013 1.294 1.295 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.218 1.221 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.169 1.193 apply_single 119 13.6 0.000 0.000 1.169 1.192 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.150 1.164 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.125 1.129 wfi_extrapolate 11 7.9 0.001 0.001 1.086 1.086 hybrid_alltoall_any 4725 16.4 0.063 0.175 0.793 1.024 make_images_data 4572 15.5 0.038 0.041 0.811 1.000 acc_transpose_blocks 9144 15.5 0.038 0.040 0.945 0.952 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.922 cp_fm_cholesky_invert 11 10.9 0.897 0.900 0.897 0.900 mp_alltoall_d11v 2130 13.8 0.774 0.835 0.774 0.835 multiply_cannon_sync_h2d 9144 15.5 0.710 0.793 0.710 0.793 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.786 0.788 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.715 0.724 qs_env_update_s_mstruct 11 6.9 0.001 0.002 0.672 0.724 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.335 0.712 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.697 0.701 mp_alltoall_z22v 1201 16.6 0.617 0.648 0.617 0.648 mp_allgather_i34 2286 14.5 0.225 0.646 0.225 0.646 mp_waitany 5200 13.7 0.456 0.589 0.456 0.589 jit_kernel_transpose 5 15.6 0.579 0.583 0.579 0.583 qs_create_task_list 11 7.9 0.001 0.002 0.540 0.568 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.361000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=568.636364, yerr=4.637255 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 763.899904E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 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 0.046 0.054 42.814 42.815 qs_mol_dyn_low 1 2.0 0.003 0.003 42.520 42.528 qs_forces 11 3.9 0.004 0.005 42.457 42.458 qs_energies 11 4.9 0.002 0.002 40.479 40.485 scf_env_do_scf 11 5.9 0.001 0.001 33.574 33.574 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.626 25.628 velocity_verlet 10 3.0 0.001 0.001 23.569 23.575 dbcsr_multiply_generic 2286 12.5 0.101 0.104 18.173 18.374 qs_scf_new_mos 108 7.5 0.001 0.001 15.877 15.975 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.876 15.974 multiply_cannon 2286 13.5 0.308 0.314 14.262 15.215 ot_scf_mini 108 9.5 0.002 0.002 14.789 14.888 multiply_cannon_loop 2286 14.5 0.346 0.352 13.005 13.948 multiply_cannon_multrec 9144 15.5 3.404 4.725 8.969 9.046 ot_mini 108 10.5 0.001 0.001 8.837 8.955 init_scf_loop 11 6.9 0.000 0.000 7.917 7.920 rebuild_ks_matrix 119 8.3 0.000 0.000 7.267 7.410 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.266 7.409 prepare_preconditioner 11 7.9 0.000 0.000 6.947 6.961 make_preconditioner 11 8.9 0.000 0.000 6.947 6.961 qs_ot_get_derivative 108 11.5 0.001 0.001 6.801 6.903 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.504 6.840 dbcsr_mm_accdrv_process 12550 15.8 4.602 6.046 5.443 6.767 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.556 6.686 init_scf_run 11 5.9 0.000 0.001 4.881 4.881 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.881 4.881 cp_fm_upper_to_full 72 14.2 3.235 4.621 3.235 4.621 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.265 4.276 calculate_rho_elec 119 8.7 0.118 0.121 4.264 4.275 sum_up_and_integrate 119 10.3 0.064 0.066 3.995 4.000 integrate_v_rspace 119 11.3 0.003 0.003 3.930 3.936 calculate_first_density_matrix 1 7.0 0.000 0.000 3.412 3.413 qs_ot_get_p 119 10.4 0.001 0.001 3.217 3.354 mp_waitall_1 94719 16.7 2.382 3.320 2.382 3.320 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.518 2.931 calculate_dm_sparse 119 9.5 0.000 0.000 2.863 2.884 dbcsr_complete_redistribute 329 12.2 0.285 0.290 1.987 2.860 pw_transfer 1439 11.6 0.068 0.069 2.686 2.693 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.588 2.595 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.674 2.526 make_m2s 4572 13.5 0.037 0.038 2.328 2.499 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.188 2.472 apply_single 119 13.6 0.000 0.000 2.188 2.472 make_images 4572 14.5 0.350 0.383 2.207 2.377 mp_alltoall_i22 627 13.8 1.452 2.291 1.452 2.291 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.438 2.280 density_rs2pw 119 9.7 0.004 0.004 2.179 2.193 fft_wrap_pw1pw2_140 487 13.2 0.617 0.620 2.180 2.189 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.268 2.180 grid_integrate_task_list 119 12.3 2.056 2.073 2.056 2.073 ot_diis_step 108 11.5 0.014 0.015 2.012 2.012 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.943 1.944 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.921 1.931 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.905 1.906 mp_sum_l 7287 12.8 1.035 1.777 1.035 1.777 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.755 1.758 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.678 1.731 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.663 1.664 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.625 1.639 fft3d_ps 1201 14.6 0.600 0.613 1.604 1.607 grid_collocate_task_list 119 9.7 1.505 1.520 1.505 1.520 cp_fm_cholesky_invert 11 10.9 1.494 1.498 1.494 1.498 rs_pw_transfer 974 11.9 0.009 0.009 1.453 1.476 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.175 1.427 wfi_extrapolate 11 7.9 0.001 0.001 1.399 1.399 potential_pw2rs 119 12.3 0.014 0.015 1.395 1.397 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.374 1.375 cp_fm_diag_elpa_base 50 14.0 1.228 1.281 1.372 1.372 make_images_data 4572 15.5 0.042 0.044 1.124 1.339 acc_transpose_blocks 9144 15.5 0.039 0.040 1.202 1.207 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.144 1.167 dbcsr_dot_sd 1205 11.9 0.326 0.329 0.799 1.149 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.091 1.112 mp_alltoall_d11v 2130 13.8 1.068 1.095 1.068 1.095 multiply_cannon_sync_h2d 9144 15.5 1.038 1.042 1.038 1.042 mp_sum_d 4133 12.0 0.636 1.027 0.636 1.027 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.980 1.016 jit_kernel_multiply 6 15.5 0.814 1.009 0.814 1.009 qs_create_task_list 11 7.9 0.001 0.002 0.933 0.950 generate_qs_task_list 11 8.9 0.368 0.388 0.932 0.948 acc_transpose_blocks_kernels 9144 16.5 0.117 0.118 0.943 0.947 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.922 0.935 mp_alltoall_z22v 1201 16.6 0.868 0.886 0.868 0.886 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.815000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=714.727273, yerr=15.094195 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.383168E+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 66218. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1543764. 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 0.013 0.028 83.347 83.347 qs_mol_dyn_low 1 2.0 0.003 0.003 82.460 83.048 qs_forces 11 3.9 0.003 0.004 82.383 82.384 qs_energies 11 4.9 0.002 0.002 79.462 79.482 scf_env_do_scf 11 5.9 0.001 0.001 69.534 69.536 scf_env_do_scf_inner_loop 99 6.5 0.041 0.108 63.855 63.855 dbcsr_multiply_generic 2055 12.4 0.108 0.111 50.591 50.879 qs_scf_new_mos 99 7.5 0.000 0.001 46.227 46.363 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.227 46.362 ot_scf_mini 99 9.5 0.002 0.002 43.886 44.012 multiply_cannon 2055 13.4 0.192 0.200 42.347 43.262 velocity_verlet 10 3.0 0.001 0.002 42.973 42.974 multiply_cannon_loop 2055 14.4 1.527 1.565 41.348 42.254 ot_mini 99 10.5 0.001 0.001 26.065 26.216 qs_ot_get_derivative 99 11.5 0.001 0.001 19.323 19.497 multiply_cannon_multrec 49320 15.4 12.504 13.212 17.438 18.054 rebuild_ks_matrix 110 8.3 0.000 0.000 14.769 14.931 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.769 14.931 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.911 13.051 mp_waitall_1 220248 16.4 10.237 11.150 10.237 11.150 multiply_cannon_sync_h2d 49320 15.4 10.411 10.969 10.411 10.969 qs_ot_get_p 110 10.4 0.001 0.001 9.564 9.710 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.291 7.856 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.214 7.711 apply_single 110 13.6 0.000 0.000 7.214 7.711 init_scf_run 11 5.9 0.000 0.001 7.330 7.330 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.330 7.330 sum_up_and_integrate 110 10.3 0.036 0.042 7.250 7.263 integrate_v_rspace 110 11.3 0.003 0.003 7.214 7.236 multiply_cannon_metrocomm3 49320 15.4 0.078 0.081 6.080 7.130 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.753 6.876 calculate_rho_elec 110 8.6 0.021 0.026 6.752 6.875 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.561 6.582 ot_diis_step 99 11.5 0.005 0.006 6.538 6.539 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.760 5.760 init_scf_loop 11 6.9 0.000 0.000 5.644 5.645 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.221 5.283 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.185 5.212 cp_fm_diag_elpa_base 48 14.0 5.172 5.201 5.183 5.210 dbcsr_mm_accdrv_process 87628 16.1 1.951 2.008 4.807 5.048 mp_sum_l 6594 12.7 3.806 4.697 3.806 4.697 rs_pw_transfer 902 11.9 0.011 0.013 3.801 4.248 wfi_extrapolate 11 7.9 0.001 0.001 4.104 4.104 make_m2s 4110 13.4 0.060 0.064 3.974 4.075 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.021 4.025 make_images 4110 14.4 0.177 0.193 3.879 3.981 density_rs2pw 110 9.6 0.004 0.005 3.480 3.958 calculate_dm_sparse 110 9.5 0.001 0.001 3.731 3.833 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.685 3.729 prepare_preconditioner 11 7.9 0.000 0.000 3.439 3.456 make_preconditioner 11 8.9 0.000 0.000 3.439 3.456 grid_integrate_task_list 110 12.3 3.257 3.404 3.257 3.404 pw_transfer 1331 11.6 0.055 0.065 3.246 3.325 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.227 3.268 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.157 3.239 multiply_cannon_metrocomm1 49320 15.4 0.061 0.063 2.210 3.131 calculate_first_density_matrix 1 7.0 0.000 0.000 3.127 3.130 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.987 3.019 potential_pw2rs 110 12.3 0.006 0.007 2.741 2.767 fft_wrap_pw1pw2_140 451 13.1 0.454 0.499 2.668 2.756 jit_kernel_multiply 13 15.9 2.581 2.648 2.581 2.648 acc_transpose_blocks 49320 15.4 0.211 0.218 2.550 2.616 mp_alltoall_d11v 2046 13.8 2.103 2.459 2.103 2.459 grid_collocate_task_list 110 9.6 2.152 2.417 2.152 2.417 fft3d_ps 1111 14.6 0.797 0.883 2.351 2.402 mp_waitany 14300 13.8 1.846 2.399 1.846 2.399 mp_sum_d 3889 11.9 1.442 2.181 1.442 2.181 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.989 2.015 qs_energies_init_hamiltonians 11 5.9 0.050 0.394 1.946 1.947 make_images_data 4110 15.4 0.042 0.046 1.801 1.938 cp_fm_cholesky_invert 11 10.9 1.907 1.911 1.907 1.911 hybrid_alltoall_any 4261 16.3 0.082 0.483 1.547 1.838 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.669 1.693 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.347000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.636364, yerr=2.837179 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 586.760192E+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 66437. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 3012787. 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 0.019 0.027 70.697 70.698 qs_mol_dyn_low 1 2.0 0.003 0.003 70.448 70.477 qs_forces 11 3.9 0.003 0.005 70.373 70.374 qs_energies 11 4.9 0.002 0.003 66.997 67.001 scf_env_do_scf 11 5.9 0.001 0.001 58.016 58.019 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 50.370 50.371 dbcsr_multiply_generic 2055 12.4 0.117 0.121 38.722 38.984 velocity_verlet 10 3.0 0.001 0.002 36.536 36.537 qs_scf_new_mos 99 7.5 0.001 0.001 33.879 34.003 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.879 34.003 multiply_cannon 2055 13.4 0.224 0.251 31.993 33.364 ot_scf_mini 99 9.5 0.003 0.003 32.221 32.332 multiply_cannon_loop 2055 14.4 0.924 0.955 30.654 31.712 ot_mini 99 10.5 0.001 0.001 19.000 19.125 multiply_cannon_multrec 24660 15.4 7.559 9.370 14.099 16.147 rebuild_ks_matrix 110 8.3 0.000 0.000 13.888 14.036 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.887 14.035 qs_ot_get_derivative 99 11.5 0.001 0.001 13.134 13.248 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.200 12.324 mp_waitall_1 176588 16.5 8.069 10.776 8.069 10.776 multiply_cannon_sync_h2d 24660 15.4 6.938 8.125 6.938 8.125 multiply_cannon_metrocomm3 24660 15.4 0.067 0.072 5.542 7.992 init_scf_loop 11 6.9 0.000 0.000 7.610 7.610 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.542 7.210 apply_single 110 13.6 0.000 0.001 6.542 7.210 dbcsr_mm_accdrv_process 52282 16.1 4.537 5.733 6.379 6.913 qs_ot_get_p 110 10.4 0.001 0.001 6.580 6.733 sum_up_and_integrate 110 10.3 0.053 0.060 6.719 6.732 integrate_v_rspace 110 11.3 0.002 0.003 6.666 6.679 init_scf_run 11 5.9 0.000 0.001 6.601 6.602 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.601 6.602 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.297 6.303 calculate_rho_elec 110 8.6 0.040 0.047 6.297 6.303 ot_diis_step 99 11.5 0.010 0.010 5.816 5.816 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.067 5.780 prepare_preconditioner 11 7.9 0.000 0.000 5.582 5.598 make_preconditioner 11 8.9 0.000 0.000 5.582 5.598 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.183 5.335 make_m2s 4110 13.4 0.057 0.060 4.247 4.717 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.661 4.684 make_images 4110 14.4 0.401 0.447 4.135 4.603 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.226 4.227 density_rs2pw 110 9.6 0.004 0.005 3.354 3.899 pw_transfer 1331 11.6 0.066 0.073 3.699 3.841 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.591 3.737 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.646 3.659 cp_fm_diag_elpa_base 48 14.0 3.597 3.613 3.643 3.656 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.654 3.656 rs_pw_transfer 902 11.9 0.012 0.014 3.041 3.609 wfi_extrapolate 11 7.9 0.001 0.001 3.582 3.582 grid_integrate_task_list 110 12.3 3.153 3.333 3.153 3.333 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.241 3.298 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.233 3.295 fft_wrap_pw1pw2_140 451 13.1 0.524 0.538 3.054 3.206 calculate_dm_sparse 110 9.5 0.001 0.001 2.966 2.995 calculate_first_density_matrix 1 7.0 0.000 0.000 2.919 2.922 make_images_data 4110 15.4 0.046 0.050 2.400 2.884 hybrid_alltoall_any 4261 16.3 0.102 0.443 2.085 2.709 fft3d_ps 1111 14.6 1.110 1.331 2.559 2.704 jit_kernel_multiply 13 16.2 1.486 2.700 1.486 2.700 cp_fm_cholesky_invert 11 10.9 2.625 2.631 2.625 2.631 mp_sum_l 6594 12.7 1.849 2.538 1.849 2.538 grid_collocate_task_list 110 9.6 2.101 2.532 2.101 2.532 potential_pw2rs 110 12.3 0.008 0.009 2.480 2.497 mp_alltoall_d11v 2046 13.8 1.763 2.147 1.763 2.147 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.972 1.998 acc_transpose_blocks 24660 15.4 0.112 0.115 1.900 1.919 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.867 1.868 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.844 1.864 mp_waitany 10164 13.8 1.262 1.785 1.262 1.785 mp_allgather_i34 2055 14.4 0.662 1.732 0.662 1.732 multiply_cannon_metrocomm4 22605 15.4 0.077 0.084 0.803 1.617 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.542 1.552 rs_pw_transfer_RS2PW_140 121 11.5 0.204 0.214 0.984 1.548 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.369 1.478 dbcsr_complete_redistribute 325 12.2 0.277 0.374 1.196 1.466 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.698000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.181818, yerr=6.712515 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 658.845696E+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 66428. 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 0.041 0.062 62.350 62.350 qs_mol_dyn_low 1 2.0 0.003 0.003 61.892 61.903 qs_forces 11 3.9 0.003 0.003 61.824 61.825 qs_energies 11 4.9 0.003 0.012 58.348 58.351 scf_env_do_scf 11 5.9 0.001 0.001 49.286 49.287 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.543 40.544 velocity_verlet 10 3.0 0.001 0.002 33.008 33.010 dbcsr_multiply_generic 2055 12.4 0.108 0.112 29.684 30.000 qs_scf_new_mos 99 7.5 0.001 0.001 25.282 25.395 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.281 25.394 ot_scf_mini 99 9.5 0.002 0.003 24.054 24.185 multiply_cannon 2055 13.4 0.296 0.332 22.766 23.989 multiply_cannon_loop 2055 14.4 0.618 0.631 21.464 22.493 ot_mini 99 10.5 0.001 0.001 13.705 13.836 rebuild_ks_matrix 110 8.3 0.000 0.000 12.768 12.915 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.018 12.768 12.914 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.015 11.142 multiply_cannon_multrec 16440 15.4 3.795 4.688 9.895 10.787 mp_waitall_1 139946 16.5 7.366 10.265 7.366 10.265 qs_ot_get_derivative 99 11.5 0.001 0.001 9.162 9.292 init_scf_loop 11 6.9 0.001 0.002 8.708 8.709 multiply_cannon_metrocomm3 16440 15.4 0.042 0.043 4.572 7.399 prepare_preconditioner 11 7.9 0.000 0.000 6.824 6.840 make_preconditioner 11 8.9 0.000 0.001 6.824 6.840 sum_up_and_integrate 110 10.3 0.061 0.062 6.642 6.656 init_scf_run 11 5.9 0.000 0.001 6.613 6.614 scf_env_initial_rho_setup 11 6.9 0.000 0.002 6.613 6.614 integrate_v_rspace 110 11.3 0.003 0.013 6.581 6.595 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.154 6.503 dbcsr_mm_accdrv_process 34862 16.1 4.922 5.544 5.954 6.141 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.098 6.106 calculate_rho_elec 110 8.6 0.059 0.059 6.097 6.106 qs_ot_get_p 110 10.4 0.001 0.001 5.503 5.678 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.088 5.556 apply_single 110 13.6 0.000 0.000 5.087 5.556 make_m2s 4110 13.4 0.049 0.051 4.644 4.997 make_images 4110 14.4 0.392 0.515 4.528 4.880 ot_diis_step 99 11.5 0.011 0.011 4.512 4.512 density_rs2pw 110 9.6 0.004 0.005 3.163 4.441 multiply_cannon_sync_h2d 16440 15.4 3.662 4.240 3.662 4.240 rs_pw_transfer 902 11.9 0.068 0.243 3.012 4.236 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.856 3.862 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.118 3.787 pw_transfer 1331 11.6 0.066 0.074 3.700 3.711 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.592 3.607 calculate_first_density_matrix 1 7.0 0.001 0.005 3.513 3.514 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.511 3.511 grid_integrate_task_list 110 12.3 3.189 3.397 3.189 3.397 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.362 3.363 fft_wrap_pw1pw2_140 451 13.1 0.639 0.649 3.091 3.103 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.026 3.082 wfi_extrapolate 11 7.9 0.001 0.001 3.007 3.007 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.966 2.977 cp_fm_diag_elpa_base 48 14.0 2.898 2.922 2.964 2.975 make_images_data 4110 15.4 0.043 0.046 2.504 2.970 calculate_dm_sparse 110 9.5 0.001 0.001 2.899 2.934 hybrid_alltoall_any 4261 16.3 0.105 0.375 2.218 2.847 mp_waitany 17072 13.8 1.414 2.764 1.414 2.764 cp_fm_cholesky_invert 11 10.9 2.743 2.749 2.743 2.749 grid_collocate_task_list 110 9.6 2.124 2.519 2.124 2.519 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.895 2.467 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.373 2.442 fft3d_ps 1111 14.6 1.092 1.104 2.411 2.424 potential_pw2rs 110 12.3 0.011 0.011 2.347 2.354 mp_sum_l 6594 12.7 1.657 2.347 1.657 2.347 mp_irecv_dv 48980 15.7 0.825 2.344 0.825 2.344 jit_kernel_multiply 10 16.4 0.638 2.297 0.638 2.297 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.179 0.931 2.156 mp_alltoall_d11v 2046 13.8 1.792 2.135 1.792 2.135 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.994 1.995 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.948 1.966 dbcsr_complete_redistribute 325 12.2 0.320 0.348 1.417 1.875 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.856 1.868 cp_fm_upper_to_full 70 14.2 1.398 1.772 1.398 1.772 mp_allgather_i34 2055 14.4 0.539 1.712 0.539 1.712 cp_fm_cholesky_decompose 22 10.9 1.581 1.601 1.581 1.601 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.364 1.469 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.971 1.436 acc_transpose_blocks 16440 15.4 0.074 0.076 1.279 1.314 rs_gather_matrices 110 12.3 0.232 0.260 0.967 1.284 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.276 1.284 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.350000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.454545, yerr=7.901250 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 735.973376E+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 66426. 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 0.052 0.116 68.021 68.022 qs_mol_dyn_low 1 2.0 0.003 0.003 66.756 66.809 qs_forces 11 3.9 0.003 0.003 66.679 66.682 qs_energies 11 4.9 0.002 0.007 63.219 63.226 scf_env_do_scf 11 5.9 0.002 0.010 54.351 54.353 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.404 42.404 velocity_verlet 10 3.0 0.001 0.002 37.537 37.539 dbcsr_multiply_generic 2055 12.4 0.115 0.121 29.924 30.164 qs_scf_new_mos 99 7.5 0.001 0.001 27.080 27.191 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.080 27.190 ot_scf_mini 99 9.5 0.003 0.003 25.446 25.532 multiply_cannon 2055 13.4 0.244 0.268 22.807 23.989 multiply_cannon_loop 2055 14.4 0.890 0.914 21.326 21.978 ot_mini 99 10.5 0.001 0.001 14.381 14.493 multiply_cannon_multrec 24660 15.4 4.212 6.753 12.754 13.891 rebuild_ks_matrix 110 8.3 0.000 0.000 12.469 12.564 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.469 12.564 init_scf_loop 11 6.9 0.001 0.003 11.901 11.904 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.998 11.084 qs_ot_get_derivative 99 11.5 0.001 0.001 10.215 10.306 prepare_preconditioner 11 7.9 0.000 0.000 10.056 10.073 make_preconditioner 11 8.9 0.001 0.002 10.056 10.073 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.264 9.735 dbcsr_mm_accdrv_process 52304 16.0 7.068 8.196 8.395 9.392 sum_up_and_integrate 110 10.3 0.068 0.073 6.695 6.717 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.686 6.693 calculate_rho_elec 110 8.6 0.078 0.081 6.685 6.693 integrate_v_rspace 110 11.3 0.003 0.003 6.627 6.648 mp_waitall_1 121746 16.5 4.436 6.583 4.436 6.583 qs_ot_get_p 110 10.4 0.001 0.001 6.001 6.125 init_scf_run 11 5.9 0.000 0.001 6.109 6.109 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.109 6.109 make_m2s 4110 13.4 0.059 0.062 5.442 5.793 make_images 4110 14.4 0.576 0.703 5.302 5.650 cp_fm_upper_to_full 70 14.2 3.357 4.863 3.357 4.863 qs_ot_p2m_diag 48 11.0 0.055 0.067 4.189 4.204 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.032 4.151 apply_single 110 13.6 0.000 0.000 4.032 4.151 ot_diis_step 99 11.5 0.011 0.011 4.126 4.126 density_rs2pw 110 9.6 0.004 0.004 3.463 4.081 pw_transfer 1331 11.6 0.065 0.075 3.891 3.930 dbcsr_complete_redistribute 325 12.2 0.415 0.462 2.681 3.837 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.785 3.828 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.696 3.697 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.486 3.571 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.506 3.552 grid_integrate_task_list 110 12.3 3.289 3.492 3.289 3.492 rs_pw_transfer 902 11.9 0.010 0.011 2.849 3.484 multiply_cannon_sync_h2d 24660 15.4 3.177 3.341 3.177 3.341 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.149 3.285 fft_wrap_pw1pw2_140 451 13.1 0.671 0.692 3.223 3.272 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.109 3.162 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.125 3.135 cp_fm_diag_elpa_base 48 14.0 2.975 3.034 3.123 3.133 hybrid_alltoall_any 4261 16.3 0.120 0.457 2.334 3.100 make_images_data 4110 15.4 0.045 0.050 2.733 3.065 wfi_extrapolate 11 7.9 0.001 0.001 3.026 3.026 calculate_dm_sparse 110 9.5 0.001 0.001 2.929 2.957 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.919 2.921 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.783 2.902 grid_collocate_task_list 110 9.6 2.318 2.865 2.318 2.865 mp_alltoall_i22 605 13.7 1.661 2.838 1.661 2.838 cp_fm_cholesky_invert 11 10.9 2.807 2.817 2.807 2.817 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.562 2.599 calculate_first_density_matrix 1 7.0 0.000 0.002 2.571 2.574 fft3d_ps 1111 14.6 1.091 1.127 2.543 2.563 potential_pw2rs 110 12.3 0.013 0.013 2.321 2.338 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.259 2.263 mp_alltoall_d11v 2046 13.8 1.808 2.066 1.808 2.066 mp_waitany 13376 13.8 1.418 2.016 1.418 2.016 rs_pw_transfer_RS2PW_140 121 11.5 0.167 0.177 1.146 1.779 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.742 1.770 cp_fm_cholesky_decompose 22 10.9 1.701 1.764 1.701 1.764 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.715 1.727 jit_kernel_multiply 8 15.8 0.994 1.714 0.994 1.714 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.598 1.698 mp_sum_l 6594 12.7 1.038 1.693 1.038 1.693 mp_allgather_i34 2055 14.4 0.576 1.660 0.576 1.660 acc_transpose_blocks 24660 15.4 0.103 0.106 1.616 1.638 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.534 1.548 multiply_cannon_metrocomm4 20550 15.4 0.059 0.063 0.859 1.462 mp_irecv_dv 62702 16.1 0.759 1.386 0.759 1.386 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=68.022000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=694.454545, yerr=9.119047 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 817.799168E+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 66424. 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 0.029 56.008 56.009 qs_mol_dyn_low 1 2.0 0.003 0.003 55.765 55.774 qs_forces 11 3.9 0.003 0.004 55.699 55.701 qs_energies 11 4.9 0.002 0.002 51.961 51.965 scf_env_do_scf 11 5.9 0.000 0.001 43.175 43.175 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.358 35.358 velocity_verlet 10 3.0 0.001 0.001 31.412 31.415 dbcsr_multiply_generic 2055 12.4 0.105 0.109 23.311 23.437 qs_scf_new_mos 99 7.5 0.001 0.001 20.219 20.280 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.218 20.280 multiply_cannon 2055 13.4 0.265 0.290 17.904 19.150 ot_scf_mini 99 9.5 0.002 0.002 18.969 19.025 multiply_cannon_loop 2055 14.4 0.322 0.334 16.557 16.748 rebuild_ks_matrix 110 8.3 0.000 0.000 11.972 12.003 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.972 12.002 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.607 10.636 ot_mini 99 10.5 0.001 0.001 10.375 10.425 multiply_cannon_multrec 8220 15.4 3.211 4.531 7.536 8.591 mp_waitall_1 103326 16.6 6.138 7.974 6.138 7.974 init_scf_loop 11 6.9 0.000 0.000 7.770 7.771 qs_ot_get_derivative 99 11.5 0.001 0.001 6.540 6.594 sum_up_and_integrate 110 10.3 0.079 0.081 6.573 6.584 integrate_v_rspace 110 11.3 0.003 0.003 6.493 6.505 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.377 6.389 calculate_rho_elec 110 8.6 0.114 0.115 6.377 6.388 prepare_preconditioner 11 7.9 0.000 0.000 6.107 6.116 make_preconditioner 11 8.9 0.000 0.000 6.107 6.116 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.679 5.758 init_scf_run 11 5.9 0.000 0.001 5.656 5.656 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.655 5.656 dbcsr_mm_accdrv_process 17442 15.9 2.893 4.080 4.194 5.110 qs_ot_get_p 110 10.4 0.001 0.001 4.659 4.713 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.167 4.624 make_m2s 4110 13.4 0.039 0.040 4.143 4.387 make_images 4110 14.4 0.634 0.691 4.012 4.257 pw_transfer 1331 11.6 0.066 0.071 3.968 3.977 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.861 3.874 ot_diis_step 99 11.5 0.012 0.012 3.815 3.815 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.729 3.769 apply_single 110 13.6 0.000 0.000 3.729 3.769 grid_integrate_task_list 110 12.3 3.369 3.521 3.369 3.521 density_rs2pw 110 9.6 0.004 0.004 3.084 3.460 fft_wrap_pw1pw2_140 451 13.1 0.837 0.849 3.326 3.343 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.325 3.328 multiply_cannon_sync_h2d 8220 15.4 2.903 3.069 2.903 3.069 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.026 3.027 cp_fm_cholesky_invert 11 10.9 2.967 2.971 2.967 2.971 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.915 2.917 calculate_first_density_matrix 1 7.0 0.000 0.000 2.778 2.779 wfi_extrapolate 11 7.9 0.001 0.001 2.777 2.777 make_images_data 4110 15.4 0.037 0.043 2.357 2.769 hybrid_alltoall_any 4261 16.3 0.199 0.860 2.288 2.678 qs_energies_init_hamiltonians 11 5.9 0.004 0.005 2.667 2.668 rs_pw_transfer 902 11.9 0.010 0.010 2.233 2.638 calculate_dm_sparse 110 9.5 0.001 0.001 2.586 2.628 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.561 2.577 grid_collocate_task_list 110 9.6 2.315 2.572 2.315 2.572 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.512 2.519 cp_fm_diag_elpa_base 48 14.0 2.455 2.478 2.510 2.518 fft3d_ps 1111 14.6 1.135 1.145 2.402 2.414 potential_pw2rs 110 12.3 0.016 0.016 2.095 2.098 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.770 2.001 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.871 1.899 mp_alltoall_d11v 2046 13.8 1.646 1.849 1.646 1.849 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.775 1.807 cp_fm_cholesky_decompose 22 10.9 1.698 1.709 1.698 1.709 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.694 1.698 qs_env_update_s_mstruct 11 6.9 0.001 0.003 1.502 1.627 mp_allgather_i34 2055 14.4 0.510 1.582 0.510 1.582 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.549 1.560 jit_kernel_multiply 9 15.8 0.987 1.533 0.987 1.533 mp_waitany 9240 13.8 1.080 1.525 1.080 1.525 dbcsr_complete_redistribute 325 12.2 0.559 0.575 1.431 1.516 qs_create_task_list 11 7.9 0.001 0.001 1.214 1.321 generate_qs_task_list 11 8.9 0.375 0.442 1.213 1.321 acc_transpose_blocks 8220 15.4 0.035 0.036 1.281 1.311 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.481 1.240 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.165 0.812 1.226 multiply_cannon_metrocomm1 8220 15.4 0.021 0.021 0.760 1.207 mp_irecv_dv 24056 15.7 0.457 1.194 0.457 1.194 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.127 1.138 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.009000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=772.636364, yerr=10.350869 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 1.345569E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 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 0.074 0.096 98.996 98.996 qs_mol_dyn_low 1 2.0 0.003 0.003 98.576 98.585 qs_forces 11 3.9 0.024 0.027 98.501 98.502 qs_energies 11 4.9 0.001 0.002 94.005 94.009 scf_env_do_scf 11 5.9 0.000 0.001 83.063 83.063 velocity_verlet 10 3.0 0.001 0.001 62.504 62.510 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 51.533 51.534 init_scf_loop 11 6.9 0.000 0.000 31.434 31.436 dbcsr_multiply_generic 2055 12.4 0.119 0.125 30.796 30.898 qs_scf_new_mos 99 7.5 0.001 0.001 29.881 29.925 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.880 29.924 prepare_preconditioner 11 7.9 0.000 0.000 29.174 29.185 make_preconditioner 11 8.9 0.000 0.000 29.174 29.185 make_full_inverse_cholesky 11 9.9 0.000 0.000 23.220 28.566 ot_scf_mini 99 9.5 0.002 0.002 28.053 28.097 multiply_cannon 2055 13.4 0.345 0.358 21.631 22.524 multiply_cannon_loop 2055 14.4 0.343 0.345 19.751 20.304 cp_fm_upper_to_full 70 14.2 12.711 18.360 12.711 18.360 rebuild_ks_matrix 110 8.3 0.000 0.000 16.157 16.194 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 16.157 16.194 ot_mini 99 10.5 0.001 0.001 15.236 15.283 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.572 14.606 dbcsr_complete_redistribute 325 12.2 1.010 1.030 7.905 11.159 qs_ot_get_derivative 99 11.5 0.001 0.001 10.671 10.711 mp_waitall_1 84994 16.7 9.410 10.426 9.410 10.426 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.672 9.921 multiply_cannon_multrec 8220 15.4 4.376 4.540 9.598 9.777 qs_rho_update_rho_low 110 7.6 0.001 0.001 9.715 9.749 calculate_rho_elec 110 8.6 0.224 0.224 9.714 9.749 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.939 9.153 sum_up_and_integrate 110 10.3 0.150 0.152 8.873 8.890 mp_alltoall_i22 605 13.7 5.515 8.775 5.515 8.775 integrate_v_rspace 110 11.3 0.003 0.003 8.723 8.740 make_m2s 4110 13.4 0.042 0.044 7.069 7.567 qs_ot_get_p 110 10.4 0.001 0.001 7.402 7.460 make_images 4110 14.4 0.876 0.921 6.877 7.375 pw_transfer 1331 11.6 0.075 0.075 6.886 6.890 cp_fm_cholesky_invert 11 10.9 6.847 6.852 6.847 6.852 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 6.768 6.773 init_scf_run 11 5.9 0.000 0.001 6.467 6.467 scf_env_initial_rho_setup 11 6.9 0.000 0.000 6.466 6.467 fft_wrap_pw1pw2_140 451 13.1 1.340 1.348 5.652 5.657 qs_ot_p2m_diag 48 11.0 0.151 0.156 5.512 5.518 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.954 5.467 apply_single 110 13.6 0.000 0.000 4.954 5.467 density_rs2pw 110 9.6 0.004 0.004 5.396 5.421 dbcsr_mm_accdrv_process 11614 15.7 3.356 3.678 5.080 5.300 make_images_data 4110 15.4 0.040 0.043 4.091 5.115 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.985 4.985 hybrid_alltoall_any 4261 16.3 0.259 0.578 3.930 4.968 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.607 4.922 fft3d_ps 1111 14.6 1.312 1.322 4.695 4.707 ot_diis_step 99 11.5 0.015 0.015 4.548 4.548 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.199 4.199 cp_fm_diag_elpa_base 48 14.0 3.612 3.863 4.197 4.197 wfi_extrapolate 11 7.9 0.001 0.001 3.992 3.992 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.275 3.989 multiply_cannon_sync_h2d 8220 15.4 3.948 3.951 3.948 3.951 grid_integrate_task_list 110 12.3 3.694 3.786 3.694 3.786 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.738 3.743 rs_pw_transfer 902 11.9 0.010 0.010 3.726 3.741 potential_pw2rs 110 12.3 0.021 0.021 3.628 3.635 calculate_dm_sparse 110 9.5 0.001 0.001 3.218 3.241 mp_alltoall_d11v 2046 13.8 3.013 3.141 3.013 3.141 mp_alltoall_z22v 1111 16.6 3.129 3.136 3.129 3.136 cp_fm_cholesky_decompose 22 10.9 3.082 3.108 3.082 3.108 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.054 3.057 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.895 2.919 grid_collocate_task_list 110 9.6 2.684 2.714 2.684 2.714 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.424 2.432 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.260 2.345 calculate_first_density_matrix 1 7.0 0.000 0.000 2.320 2.320 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.187 2.191 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.093 2.156 copy_dbcsr_to_fm 151 11.3 0.003 0.003 2.039 2.065 make_images_sizes 4110 15.4 0.005 0.005 0.999 2.038 mp_alltoall_i44 4110 16.4 0.994 2.033 0.994 2.033 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=98.996000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1210.272727, yerr=53.838295 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 630.870016E+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 57910. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2361102. 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 0.014 0.033 230.410 230.411 qs_mol_dyn_low 1 2.0 0.003 0.003 229.943 229.956 qs_forces 11 3.9 0.004 0.005 228.897 228.898 qs_energies 11 4.9 0.002 0.003 222.931 222.945 scf_env_do_scf 11 5.9 0.001 0.005 204.441 204.445 scf_env_do_scf_inner_loop 117 6.6 0.002 0.007 176.922 176.925 velocity_verlet 10 3.0 0.001 0.001 139.897 139.908 qs_scf_new_mos 117 7.6 0.001 0.001 131.243 131.541 qs_scf_loop_do_ot 117 8.6 0.001 0.001 131.243 131.540 dbcsr_multiply_generic 2507 12.6 0.181 0.188 126.399 126.915 ot_scf_mini 117 9.6 0.003 0.003 124.482 124.809 multiply_cannon 2507 13.6 0.240 0.253 101.477 102.912 multiply_cannon_loop 2507 14.6 2.032 2.120 98.729 100.126 ot_mini 117 10.6 0.001 0.001 67.374 67.637 multiply_cannon_multrec 60168 15.6 33.477 35.997 41.562 43.667 qs_ot_get_derivative 117 11.6 0.001 0.001 42.623 42.908 rebuild_ks_matrix 128 8.3 0.001 0.001 36.764 37.016 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 36.763 37.016 mp_waitall_1 267128 16.5 31.105 35.236 31.105 35.236 qs_ot_get_p 128 10.4 0.001 0.001 34.931 35.230 qs_ks_update_qs_env 128 7.6 0.001 0.001 32.987 33.234 multiply_cannon_sync_h2d 60168 15.6 27.659 30.241 27.659 30.241 qs_ot_p2m_diag 83 11.4 0.079 0.091 27.627 27.713 init_scf_loop 11 6.9 0.000 0.001 27.414 27.416 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.363 25.208 apply_single 128 13.6 0.001 0.001 24.363 25.207 cp_dbcsr_syevd 83 12.4 0.005 0.005 25.182 25.185 ot_diis_step 117 11.6 0.008 0.008 24.533 24.535 prepare_preconditioner 11 7.9 0.000 0.000 22.486 22.534 make_preconditioner 11 8.9 0.000 0.000 22.486 22.534 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.653 21.816 cp_fm_diag_elpa 83 13.4 0.000 0.000 21.336 21.369 cp_fm_diag_elpa_base 83 14.4 21.212 21.258 21.331 21.363 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.456 19.626 qs_rho_update_rho_low 128 7.7 0.001 0.001 17.161 17.247 calculate_rho_elec 128 8.7 0.046 0.065 17.160 17.247 sum_up_and_integrate 128 10.3 0.089 0.106 16.961 16.977 multiply_cannon_metrocomm3 60168 15.6 0.113 0.117 14.985 16.903 integrate_v_rspace 128 11.3 0.004 0.004 16.871 16.889 make_m2s 5014 13.6 0.104 0.112 16.303 16.670 make_images 5014 14.6 0.400 0.418 16.122 16.503 init_scf_run 11 5.9 0.000 0.001 14.091 14.092 scf_env_initial_rho_setup 11 6.9 0.000 0.001 14.091 14.092 density_rs2pw 128 9.7 0.006 0.007 10.336 13.602 cp_fm_cholesky_invert 11 10.9 13.219 13.230 13.219 13.230 rs_pw_transfer 1046 11.9 0.016 0.018 9.764 13.177 wfi_extrapolate 11 7.9 0.001 0.001 10.163 10.163 pw_transfer 1547 11.6 0.075 0.093 9.636 9.942 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 9.432 9.741 make_images_data 5014 15.6 0.065 0.071 8.083 9.109 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.823 9.002 calculate_dm_sparse 128 9.5 0.001 0.001 8.610 8.722 mp_sum_l 7950 12.9 7.344 8.551 7.344 8.551 multiply_cannon_metrocomm1 60168 15.6 0.089 0.093 6.484 8.494 hybrid_alltoall_any 5200 16.5 0.289 2.260 6.853 8.389 dbcsr_mm_accdrv_process 124484 16.2 2.984 3.283 7.632 8.346 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.972 8.049 fft3d_ps 1291 14.7 2.155 2.798 7.227 7.693 grid_integrate_task_list 128 12.3 7.003 7.678 7.003 7.678 fft_wrap_pw1pw2_140 523 13.2 1.245 1.305 7.313 7.625 potential_pw2rs 128 12.3 0.009 0.011 7.100 7.235 cp_fm_cholesky_decompose 22 10.9 6.939 6.956 6.939 6.956 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.887 6.899 mp_alltoall_d11v 2415 14.1 4.982 6.641 4.982 6.641 mp_waitany 16020 13.9 2.846 6.169 2.846 6.169 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.935 6.022 grid_collocate_task_list 128 9.7 4.694 5.972 4.694 5.972 rs_pw_transfer_RS2PW_140 139 11.5 0.279 0.301 2.198 5.477 make_images_sizes 5014 15.6 0.006 0.007 2.804 4.899 mp_alltoall_i44 5014 16.6 2.797 4.893 2.797 4.893 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=230.411000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.636364, yerr=6.581969 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 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 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 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 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 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 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 2.200017E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 6024768 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.7 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 831.709184E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2427840 MPI messages size (bytes): total size 4.132587E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.702166E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71532 2339635200 32768 < size <= 131072 729952 56049532928 131072 < size <= 4194304 1387568 1410045313024 4194304 < size <= 16777216 155760 1473827979536 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57641. MP_Allreduce 11185 956. MP_Sync 88 MP_Alltoall 1983 4973905. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.040 228.105 228.106 qs_mol_dyn_low 1 2.0 0.003 0.003 227.653 227.694 qs_forces 11 3.9 0.004 0.004 227.535 227.540 qs_energies 11 4.9 0.009 0.055 220.337 220.360 scf_env_do_scf 11 5.9 0.001 0.001 202.437 202.447 scf_env_do_scf_inner_loop 118 6.6 0.004 0.008 158.991 158.994 velocity_verlet 10 3.0 0.001 0.002 144.406 144.409 qs_scf_new_mos 118 7.6 0.001 0.001 113.814 114.341 qs_scf_loop_do_ot 118 8.6 0.001 0.001 113.814 114.340 dbcsr_multiply_generic 2529 12.6 0.194 0.199 106.249 109.335 ot_scf_mini 118 9.6 0.004 0.004 108.630 109.220 multiply_cannon 2529 13.6 0.480 0.530 80.999 87.984 multiply_cannon_loop 2529 14.6 1.258 1.298 76.763 84.400 ot_mini 118 10.6 0.001 0.001 57.286 57.823 mp_waitall_1 216598 16.6 30.411 51.762 30.411 51.762 init_scf_loop 11 6.9 0.000 0.000 43.323 43.326 multiply_cannon_metrocomm3 30348 15.6 0.092 0.100 18.442 39.951 prepare_preconditioner 11 7.9 0.000 0.000 38.606 38.658 make_preconditioner 11 8.9 0.000 0.000 38.606 38.658 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.147 37.818 rebuild_ks_matrix 129 8.3 0.001 0.001 36.295 36.922 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 36.295 36.921 multiply_cannon_multrec 30348 15.6 21.652 26.618 31.321 36.498 qs_ot_get_derivative 118 11.6 0.001 0.002 34.209 34.794 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.677 33.248 qs_ot_get_p 129 10.4 0.001 0.001 31.872 32.559 qs_ot_p2m_diag 84 11.4 0.190 0.218 25.871 26.017 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.549 24.427 apply_single 129 13.6 0.001 0.001 22.549 24.427 cp_dbcsr_syevd 84 12.4 0.006 0.006 24.233 24.236 cp_fm_cholesky_invert 11 10.9 22.781 22.800 22.781 22.800 ot_diis_step 118 11.6 0.015 0.015 22.718 22.720 multiply_cannon_sync_h2d 30348 15.6 19.044 21.198 19.044 21.198 cp_fm_diag_elpa 84 13.4 0.000 0.000 19.989 20.014 cp_fm_diag_elpa_base 84 14.4 19.636 19.828 19.982 20.009 make_m2s 5058 13.6 0.090 0.095 16.589 18.126 make_images 5058 14.6 1.167 1.356 16.374 17.910 sum_up_and_integrate 129 10.3 0.117 0.135 17.194 17.230 integrate_v_rspace 129 11.3 0.003 0.004 17.077 17.117 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.914 17.022 calculate_rho_elec 129 8.7 0.088 0.105 16.913 17.022 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 14.173 14.606 init_scf_run 11 5.9 0.000 0.001 12.805 12.807 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.805 12.807 density_rs2pw 129 9.7 0.007 0.007 9.908 12.317 rs_pw_transfer 1054 12.0 0.014 0.017 9.446 11.957 make_images_data 5058 15.6 0.063 0.074 9.829 11.836 multiply_cannon_metrocomm4 27819 15.6 0.104 0.121 3.815 11.296 mp_irecv_dv 70084 16.3 3.616 10.901 3.616 10.901 cp_fm_cholesky_decompose 22 10.9 10.628 10.709 10.628 10.709 hybrid_alltoall_any 5245 16.5 0.346 1.528 8.292 10.665 pw_transfer 1559 11.6 0.088 0.106 9.931 10.009 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 9.704 9.784 dbcsr_mm_accdrv_process 62758 16.2 4.529 5.319 9.153 9.736 wfi_extrapolate 11 7.9 0.001 0.001 9.434 9.434 mp_sum_l 8016 12.9 6.882 9.207 6.882 9.207 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.376 8.140 fft_wrap_pw1pw2_140 527 13.2 1.343 1.368 7.869 7.955 grid_integrate_task_list 129 12.3 7.164 7.791 7.164 7.791 potential_pw2rs 129 12.3 0.015 0.018 7.282 7.325 fft3d_ps 1301 14.7 2.831 3.003 6.990 7.060 calculate_dm_sparse 129 9.5 0.001 0.001 6.868 7.023 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.417 6.427 qs_ot_get_orbitals 118 10.6 0.001 0.001 6.187 6.287 mp_allgather_i34 2529 14.6 2.787 6.159 2.787 6.159 grid_collocate_task_list 129 9.7 4.823 5.980 4.823 5.980 mp_alltoall_d11v 2429 14.1 5.035 5.914 5.035 5.914 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.389 5.576 mp_waitany 11836 13.9 2.844 5.365 2.844 5.365 mp_sum_d 4509 12.1 3.424 5.025 3.424 5.025 dbcsr_complete_redistribute 397 12.7 0.785 0.889 3.930 4.991 rs_pw_transfer_RS2PW_140 140 11.5 0.353 0.375 2.188 4.602 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=228.106000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=793.181818, yerr=1.991718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+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 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 947.134464E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11057 1000. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.074 181.754 181.755 qs_mol_dyn_low 1 2.0 0.011 0.070 180.775 180.789 qs_forces 11 3.9 0.031 0.050 180.667 180.677 qs_energies 11 4.9 0.013 0.028 173.886 173.932 scf_env_do_scf 11 5.9 0.001 0.002 157.243 157.244 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 121.154 121.155 velocity_verlet 10 3.0 0.001 0.002 114.962 114.964 dbcsr_multiply_generic 2507 12.6 0.180 0.185 83.559 84.594 qs_scf_new_mos 117 7.6 0.001 0.001 82.243 82.618 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.242 82.617 ot_scf_mini 117 9.6 0.004 0.004 78.065 78.466 multiply_cannon 2507 13.6 0.497 0.518 62.767 67.377 multiply_cannon_loop 2507 14.6 0.862 0.896 59.813 62.467 ot_mini 117 10.6 0.001 0.001 43.675 44.085 mp_waitall_1 170520 16.6 26.191 36.091 26.191 36.091 init_scf_loop 11 6.9 0.001 0.004 35.980 35.981 rebuild_ks_matrix 128 8.3 0.001 0.001 31.223 31.703 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 31.223 31.703 prepare_preconditioner 11 7.9 0.000 0.001 31.554 31.597 make_preconditioner 11 8.9 0.000 0.002 31.554 31.597 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.092 30.579 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.091 28.531 multiply_cannon_multrec 20056 15.6 13.341 16.687 22.332 25.552 multiply_cannon_metrocomm3 20056 15.6 0.058 0.061 15.880 25.539 qs_ot_get_derivative 117 11.6 0.002 0.002 23.699 24.106 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.336 21.398 apply_single 128 13.6 0.001 0.001 20.336 21.397 qs_ot_get_p 128 10.4 0.001 0.001 19.999 20.583 ot_diis_step 117 11.6 0.018 0.018 19.851 19.852 make_m2s 5014 13.6 0.080 0.084 15.309 16.033 multiply_cannon_sync_h2d 20056 15.6 14.080 15.962 14.080 15.962 make_images 5014 14.6 1.170 1.265 15.073 15.795 qs_ot_p2m_diag 83 11.4 0.317 0.477 15.337 15.345 sum_up_and_integrate 128 10.3 0.133 0.144 14.982 15.016 integrate_v_rspace 128 11.3 0.004 0.004 14.849 14.874 cp_fm_cholesky_invert 11 10.9 14.838 14.847 14.838 14.847 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.662 14.702 calculate_rho_elec 128 8.7 0.130 0.145 14.662 14.701 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.259 14.313 init_scf_run 11 5.9 0.000 0.001 11.384 11.385 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.384 11.384 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.934 10.946 cp_fm_diag_elpa_base 83 14.4 10.505 10.659 10.930 10.942 make_images_data 5014 15.6 0.058 0.065 9.548 10.680 hybrid_alltoall_any 5200 16.5 0.436 1.990 8.194 10.020 density_rs2pw 128 9.7 0.006 0.006 7.449 9.749 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.120 9.417 multiply_cannon_metrocomm4 17549 15.6 0.061 0.071 3.494 9.335 pw_transfer 1547 11.6 0.086 0.105 9.163 9.269 mp_irecv_dv 50230 16.2 3.372 9.084 3.372 9.084 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.938 9.050 dbcsr_mm_accdrv_process 41502 16.2 4.515 5.499 8.452 8.874 fft_wrap_pw1pw2_140 523 13.2 1.420 1.445 7.803 7.923 rs_pw_transfer 1046 11.9 0.013 0.015 5.562 7.842 grid_integrate_task_list 128 12.3 7.283 7.787 7.283 7.787 wfi_extrapolate 11 7.9 0.001 0.001 7.597 7.597 cp_fm_cholesky_decompose 22 10.9 7.501 7.525 7.501 7.525 cp_fm_upper_to_full 105 14.8 5.731 7.334 5.731 7.334 dbcsr_complete_redistribute 395 12.7 1.164 1.197 4.645 6.405 calculate_dm_sparse 128 9.5 0.001 0.001 6.290 6.364 fft3d_ps 1291 14.7 2.743 2.972 6.051 6.145 grid_collocate_task_list 128 9.7 4.967 5.933 4.967 5.933 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.836 5.775 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.688 5.693 mp_sum_l 7950 12.9 3.677 5.280 3.677 5.280 copy_fm_to_dbcsr 209 11.7 0.002 0.003 3.447 5.210 mp_alltoall_d11v 2415 14.1 4.417 5.009 4.417 5.009 potential_pw2rs 128 12.3 0.021 0.023 4.986 5.005 mp_waitany 11748 13.9 2.555 4.903 2.555 4.903 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.771 4.882 mp_allgather_i34 2507 14.6 1.471 4.600 1.471 4.600 rs_pw_transfer_RS2PW_140 139 11.5 0.328 0.349 2.074 4.342 transfer_fm_to_dbcsr 11 9.9 0.019 0.029 2.442 4.195 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.106 4.139 qs_energies_init_hamiltonians 11 5.9 0.003 0.011 3.904 3.941 mp_alltoall_i22 716 14.1 2.046 3.935 2.046 3.935 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.643 3.686 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=181.755000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=897.727273, yerr=10.566593 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/18/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 4.353788E+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 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.166156E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58354. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.058 0.068 191.500 191.501 qs_mol_dyn_low 1 2.0 0.003 0.004 190.583 190.597 qs_forces 11 3.9 0.028 0.034 190.460 190.473 qs_energies 11 4.9 0.004 0.020 183.161 183.175 scf_env_do_scf 11 5.9 0.001 0.015 165.624 165.631 velocity_verlet 10 3.0 0.001 0.002 125.798 125.800 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 118.956 118.958 qs_scf_new_mos 117 7.6 0.001 0.001 80.919 81.242 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.918 81.241 dbcsr_multiply_generic 2507 12.6 0.187 0.193 80.470 81.000 ot_scf_mini 117 9.6 0.003 0.003 76.424 76.715 multiply_cannon 2507 13.6 0.552 0.589 55.594 58.250 multiply_cannon_loop 2507 14.6 1.187 1.208 51.794 53.099 init_scf_loop 11 6.9 0.001 0.005 46.532 46.533 ot_mini 117 10.6 0.001 0.001 42.638 42.937 prepare_preconditioner 11 7.9 0.000 0.000 42.351 42.382 make_preconditioner 11 8.9 0.000 0.001 42.351 42.382 make_full_inverse_cholesky 11 9.9 0.012 0.023 36.022 40.969 multiply_cannon_multrec 30084 15.6 14.099 18.936 26.320 30.874 rebuild_ks_matrix 128 8.3 0.001 0.001 30.312 30.525 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 30.311 30.524 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.284 27.476 mp_waitall_1 147882 16.7 17.595 26.997 17.595 26.997 qs_ot_get_derivative 117 11.6 0.002 0.002 22.913 23.195 make_m2s 5014 13.6 0.095 0.099 20.435 21.592 make_images 5014 14.6 1.953 2.220 20.128 21.284 qs_ot_get_p 128 10.4 0.001 0.001 19.904 20.242 ot_diis_step 117 11.6 0.018 0.020 19.597 19.599 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.141 19.578 apply_single 128 13.6 0.001 0.001 19.141 19.578 cp_fm_upper_to_full 105 14.8 11.368 16.725 11.368 16.725 cp_fm_cholesky_invert 11 10.9 16.346 16.355 16.346 16.355 qs_ot_p2m_diag 83 11.4 0.342 0.389 15.800 15.852 multiply_cannon_metrocomm3 30084 15.6 0.048 0.050 6.633 15.070 sum_up_and_integrate 128 10.3 0.139 0.152 14.951 14.979 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.858 14.896 calculate_rho_elec 128 8.7 0.173 0.189 14.857 14.896 integrate_v_rspace 128 11.3 0.004 0.004 14.811 14.843 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.495 14.496 dbcsr_complete_redistribute 395 12.7 1.503 1.621 8.963 12.678 multiply_cannon_sync_h2d 30084 15.6 11.676 12.636 11.676 12.636 make_images_data 5014 15.6 0.061 0.065 11.025 12.450 dbcsr_mm_accdrv_process 62264 16.2 7.314 8.300 11.791 12.322 hybrid_alltoall_any 5200 16.5 0.525 2.225 9.922 11.720 init_scf_run 11 5.9 0.000 0.001 11.565 11.566 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.565 11.566 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.371 11.386 cp_fm_diag_elpa_base 83 14.4 10.405 10.724 11.364 11.377 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.573 11.298 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.306 9.918 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.480 9.702 pw_transfer 1547 11.6 0.087 0.102 9.396 9.481 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.168 9.260 mp_alltoall_i22 716 14.1 5.525 9.175 5.525 9.175 density_rs2pw 128 9.7 0.006 0.007 7.183 8.676 fft_wrap_pw1pw2_140 523 13.2 1.566 1.595 8.113 8.214 grid_integrate_task_list 128 12.3 7.487 8.051 7.487 8.051 wfi_extrapolate 11 7.9 0.001 0.001 7.802 7.802 cp_fm_cholesky_decompose 22 10.9 7.578 7.663 7.578 7.663 multiply_cannon_metrocomm4 25070 15.6 0.079 0.090 2.755 7.372 mp_irecv_dv 76098 16.2 2.608 7.097 2.608 7.097 calculate_dm_sparse 128 9.5 0.001 0.001 6.714 6.783 rs_pw_transfer 1046 11.9 0.013 0.016 4.824 6.410 fft3d_ps 1291 14.7 2.835 2.915 6.083 6.138 mp_alltoall_d11v 2415 14.1 5.063 6.126 5.063 6.126 grid_collocate_task_list 128 9.7 5.147 5.869 5.147 5.869 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.577 5.641 potential_pw2rs 128 12.3 0.024 0.024 4.690 4.709 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.523 4.586 qs_energies_init_hamiltonians 11 5.9 0.005 0.014 4.503 4.506 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.416 4.498 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.255 4.301 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=191.501000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1097.000000, yerr=21.687868 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 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 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 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 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.541583E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108613552 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3983 58480. MP_Allreduce 11035 1168. MP_Sync 86 MP_Alltoall 1700 18828162. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.031 177.795 177.796 qs_mol_dyn_low 1 2.0 0.003 0.003 177.329 177.395 qs_forces 11 3.9 0.004 0.005 176.978 176.986 qs_energies 11 4.9 0.002 0.002 169.316 169.325 scf_env_do_scf 11 5.9 0.017 0.264 151.005 151.009 velocity_verlet 10 3.0 0.004 0.026 115.147 115.164 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 114.168 114.169 dbcsr_multiply_generic 2485 12.5 0.180 0.193 75.652 76.254 qs_scf_new_mos 116 7.6 0.001 0.001 75.629 75.720 qs_scf_loop_do_ot 116 8.6 0.001 0.001 75.628 75.720 ot_scf_mini 116 9.6 0.003 0.004 71.217 71.312 multiply_cannon 2485 13.5 0.579 0.628 55.662 60.776 multiply_cannon_loop 2485 14.5 0.441 0.453 50.827 51.821 ot_mini 116 10.6 0.001 0.001 40.062 40.166 init_scf_loop 11 6.9 0.000 0.000 36.425 36.427 mp_waitall_1 124680 16.7 26.454 34.093 26.454 34.093 prepare_preconditioner 11 7.9 0.000 0.000 32.465 32.502 make_preconditioner 11 8.9 0.000 0.000 32.465 32.502 make_full_inverse_cholesky 11 9.9 0.016 0.028 30.335 30.621 rebuild_ks_matrix 127 8.3 0.001 0.001 29.949 30.050 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 29.948 30.050 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.108 27.199 multiply_cannon_multrec 9940 15.5 10.377 15.417 17.992 21.859 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 12.802 20.164 qs_ot_get_derivative 116 11.6 0.002 0.002 20.019 20.110 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.576 20.006 apply_single 127 13.6 0.001 0.001 19.576 20.006 ot_diis_step 116 11.6 0.019 0.020 19.968 19.968 make_m2s 4970 13.5 0.066 0.071 16.250 18.704 cp_fm_cholesky_invert 11 10.9 18.694 18.700 18.694 18.700 make_images 4970 14.5 2.276 2.626 15.943 18.398 qs_ot_get_p 127 10.4 0.001 0.001 18.085 18.230 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.355 15.403 calculate_rho_elec 127 8.7 0.254 0.265 15.354 15.403 sum_up_and_integrate 127 10.3 0.180 0.189 15.019 15.060 integrate_v_rspace 127 11.3 0.004 0.004 14.839 14.888 qs_ot_p2m_diag 82 11.4 0.489 0.494 14.272 14.289 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.123 13.125 make_images_data 4970 15.5 0.050 0.059 9.875 12.510 hybrid_alltoall_any 5155 16.4 0.784 3.602 9.694 12.137 multiply_cannon_sync_h2d 9940 15.5 11.436 12.004 11.436 12.004 init_scf_run 11 5.9 0.000 0.001 11.447 11.449 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.447 11.448 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.082 10.092 cp_fm_diag_elpa_base 82 14.4 9.850 9.925 10.079 10.090 pw_transfer 1535 11.6 0.086 0.096 9.593 9.618 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.370 9.402 fft_wrap_pw1pw2_140 519 13.2 1.909 1.953 8.263 8.297 cp_fm_cholesky_decompose 22 10.9 8.029 8.143 8.029 8.143 grid_integrate_task_list 127 12.3 7.694 8.099 7.694 8.099 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.985 8.059 mp_allgather_i34 2485 14.5 3.041 7.993 3.041 7.993 dbcsr_mm_accdrv_process 20590 16.1 2.635 3.422 7.249 7.877 density_rs2pw 127 9.7 0.006 0.006 6.921 7.822 wfi_extrapolate 11 7.9 0.001 0.001 7.657 7.657 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.457 7.346 calculate_dm_sparse 127 9.5 0.001 0.001 6.405 6.490 grid_collocate_task_list 127 9.7 5.426 6.123 5.426 6.123 fft3d_ps 1281 14.7 2.751 2.838 5.859 5.902 mp_alltoall_d11v 2401 14.1 5.178 5.875 5.178 5.875 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.715 5.744 dbcsr_complete_redistribute 393 12.7 2.102 2.159 5.259 5.690 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 5.306 5.307 multiply_cannon_metrocomm4 7455 15.5 0.023 0.026 1.862 5.128 rs_pw_transfer 1038 11.9 0.013 0.021 4.156 5.081 mp_irecv_dv 28618 15.9 1.827 5.044 1.827 5.044 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.616 4.651 potential_pw2rs 127 12.3 0.027 0.027 4.461 4.476 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.841 3.932 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.554 3.852 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.758 3.802 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.740 3.752 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.409 3.735 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.565 3.615 calculate_first_density_matrix 1 7.0 0.000 0.000 3.567 3.568 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=177.796000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1445.818182, yerr=56.547355 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 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 94.243758E+12 0.0% 0.0% 100.0% flops max/rank 11.787404E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547744 0.0% 0.0% 100.0% number of processed stacks 1982128 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.0 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 3.110474E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101360 MPI messages size (bytes): total size 1.145137E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.297725E+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 45848 35601252352 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58564. MP_Allreduce 11184 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.039 304.752 304.753 qs_mol_dyn_low 1 2.0 0.003 0.003 303.770 303.786 qs_forces 11 3.9 0.004 0.005 303.676 303.679 qs_energies 11 4.9 0.002 0.002 294.266 294.276 scf_env_do_scf 11 5.9 0.001 0.006 271.433 271.447 velocity_verlet 10 3.0 0.001 0.001 218.468 218.476 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 144.928 144.930 init_scf_loop 11 6.9 0.000 0.000 126.251 126.258 prepare_preconditioner 11 7.9 0.000 0.000 121.134 121.158 make_preconditioner 11 8.9 0.000 0.000 121.134 121.158 make_full_inverse_cholesky 11 9.9 0.036 0.039 96.702 118.275 qs_scf_new_mos 118 7.6 0.001 0.001 93.382 93.457 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.381 93.456 ot_scf_mini 118 9.6 0.004 0.004 88.510 88.549 dbcsr_multiply_generic 2534 12.6 0.226 0.237 85.426 86.085 cp_fm_upper_to_full 106 14.8 53.565 77.034 53.565 77.034 multiply_cannon 2534 13.6 0.710 0.802 59.976 60.798 multiply_cannon_loop 2534 14.6 0.479 0.486 56.112 57.742 ot_mini 118 10.6 0.001 0.001 45.710 45.767 dbcsr_complete_redistribute 397 12.7 3.973 4.022 30.088 43.514 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.732 40.212 rebuild_ks_matrix 129 8.3 0.001 0.001 38.848 38.926 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 38.847 38.925 transfer_fm_to_dbcsr 11 9.9 0.026 0.031 24.386 37.707 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.807 35.877 mp_alltoall_i22 720 14.1 22.237 35.851 22.237 35.851 cp_fm_cholesky_invert 11 10.9 33.918 33.924 33.918 33.924 mp_waitall_1 104780 16.8 29.393 33.368 29.393 33.368 qs_ot_get_p 129 10.4 0.001 0.001 26.947 27.032 qs_ot_get_derivative 118 11.6 0.002 0.002 24.966 25.000 qs_ot_p2m_diag 84 11.4 0.889 0.894 22.691 22.719 make_m2s 5068 13.6 0.075 0.077 20.671 21.784 make_images 5068 14.6 3.781 3.944 20.186 21.300 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.929 20.931 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.884 20.898 calculate_rho_elec 129 8.7 0.483 0.483 20.883 20.898 multiply_cannon_metrocomm3 10136 15.6 0.023 0.026 19.306 20.850 ot_diis_step 118 11.6 0.022 0.023 20.704 20.705 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.886 20.142 apply_single 129 13.6 0.001 0.001 19.886 20.142 sum_up_and_integrate 129 10.3 0.324 0.326 19.889 19.964 integrate_v_rspace 129 11.3 0.004 0.004 19.565 19.641 multiply_cannon_multrec 10136 15.6 10.669 12.534 18.413 18.591 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.652 17.653 cp_fm_diag_elpa_base 84 14.4 13.153 14.851 17.648 17.648 multiply_cannon_sync_h2d 10136 15.6 15.775 15.801 15.775 15.801 hybrid_alltoall_any 5255 16.5 1.311 3.067 11.358 13.373 make_images_data 5068 15.6 0.058 0.062 11.289 13.178 init_scf_run 11 5.9 0.000 0.001 12.752 12.752 scf_env_initial_rho_setup 11 6.9 0.000 0.000 12.751 12.752 pw_transfer 1559 11.6 0.093 0.095 11.684 11.694 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 11.445 11.454 fft_wrap_pw1pw2_140 527 13.2 3.102 3.125 10.141 10.154 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.748 9.777 mp_alltoall_d11v 2429 14.1 8.531 9.523 8.531 9.523 dbcsr_mm_accdrv_process 20954 16.1 3.871 5.667 7.501 9.397 wfi_extrapolate 11 7.9 0.001 0.001 9.197 9.197 cp_fm_cholesky_decompose 22 10.9 9.088 9.147 9.088 9.147 grid_integrate_task_list 129 12.3 8.635 8.831 8.635 8.831 density_rs2pw 129 9.7 0.005 0.005 8.207 8.369 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.018 8.020 calculate_dm_sparse 129 9.5 0.001 0.001 6.813 6.879 grid_collocate_task_list 129 9.7 6.479 6.555 6.479 6.555 fft3d_ps 1301 14.7 2.867 2.879 6.479 6.500 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.391 6.440 rs_scatter_matrices 140 9.7 3.733 4.639 6.169 6.405 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.068 6.152 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=304.753000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2811.363636, yerr=156.884597 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 1.259643E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 8765829. 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 0.024 0.081 85.986 85.987 qs_energies 1 2.0 0.000 0.000 85.408 85.438 ls_scf 1 3.0 0.004 0.029 84.324 84.337 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.974 73.221 multiply_cannon 111 7.7 0.017 0.020 55.998 57.364 multiply_cannon_loop 111 8.7 0.209 0.220 52.591 53.947 ls_scf_main 1 4.0 0.001 0.008 52.598 52.605 density_matrix_trs4 2 5.0 0.002 0.003 46.957 47.042 ls_scf_init_scf 1 4.0 0.000 0.001 28.631 28.669 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.257 27.312 mp_waitall_1 11031 10.9 22.522 26.175 22.522 26.175 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.119 25.137 multiply_cannon_multrec 2664 9.7 8.178 8.795 15.614 17.332 multiply_cannon_sync_h2d 2664 9.7 13.675 15.250 13.675 15.250 make_m2s 222 7.7 0.008 0.011 13.243 13.649 make_images 222 8.7 0.099 0.109 13.221 13.630 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.655 11.978 make_images_data 222 9.7 0.004 0.005 7.794 8.349 hybrid_alltoall_any 227 10.6 0.216 1.830 6.677 8.272 dbcsr_mm_accdrv_process 4760 10.4 0.505 0.641 7.053 8.122 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.405 7.742 dbcsr_mm_accdrv_process_sort 4760 11.4 6.330 7.220 6.330 7.220 calculate_norms 4752 9.8 5.541 6.136 5.541 6.136 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.074 5.275 mp_sum_l 887 5.1 3.222 5.007 3.222 5.007 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.343 3.735 make_images_sizes 222 9.7 0.000 0.000 0.793 3.683 mp_alltoall_i44 222 10.7 0.793 3.682 0.793 3.682 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.043 3.268 arnoldi_extremal 4 6.8 0.000 0.000 3.247 3.268 arnoldi_normal_ev 4 7.8 0.001 0.002 3.247 3.268 mp_irecv_dv 6231 10.9 2.026 3.237 2.026 3.237 build_subspace 16 8.4 0.009 0.012 3.149 3.151 ls_scf_post 1 4.0 0.001 0.017 3.091 3.105 ls_scf_store_result 1 5.0 0.000 0.000 2.908 2.951 dbcsr_special_finalize 555 9.7 0.005 0.006 2.306 2.740 dbcsr_merge_single_wm 555 10.7 0.456 0.618 2.298 2.731 make_images_pack 222 9.7 2.205 2.626 2.207 2.628 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.325 2.580 dbcsr_sort_data 658 11.4 2.102 2.488 2.102 2.488 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.476 2.070 2.478 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.360 2.453 buffer_matrices_ensure_size 222 8.7 1.747 2.109 1.747 2.109 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.988 2.031 rebuild_ks_matrix 3 7.3 0.000 0.000 1.978 2.021 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.081 1.978 2.021 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.987000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 2.111377E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 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 0.044 0.062 92.197 92.198 qs_energies 1 2.0 0.000 0.000 91.634 91.652 ls_scf 1 3.0 0.047 0.373 90.000 90.018 dbcsr_multiply_generic 111 6.7 0.016 0.024 75.377 75.746 multiply_cannon 111 7.7 0.029 0.040 53.396 57.004 ls_scf_main 1 4.0 0.000 0.010 55.327 55.379 multiply_cannon_loop 111 8.7 0.116 0.123 50.097 53.239 density_matrix_trs4 2 5.0 0.012 0.140 49.205 49.387 ls_scf_init_scf 1 4.0 0.001 0.002 31.077 31.078 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.783 29.887 mp_waitall_1 9105 10.9 21.064 29.310 21.064 29.310 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 27.374 27.383 multiply_cannon_multrec 1332 9.7 13.214 16.605 22.545 27.030 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.694 20.070 make_m2s 222 7.7 0.006 0.009 15.260 15.865 make_images 222 8.7 1.578 1.924 15.230 15.837 dbcsr_mm_accdrv_process 4041 10.4 0.285 0.455 8.934 10.559 dbcsr_mm_accdrv_process_sort 4041 11.4 8.525 10.104 8.525 10.104 make_images_data 222 9.7 0.004 0.005 8.777 9.724 hybrid_alltoall_any 227 10.6 0.522 2.430 8.259 9.247 mp_sum_l 887 5.1 5.431 8.532 5.431 8.532 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.215 7.787 mp_irecv_dv 3311 11.0 3.196 7.738 3.196 7.738 calculate_norms 2376 9.8 6.006 6.770 6.006 6.770 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.241 6.766 multiply_cannon_sync_h2d 1332 9.7 4.794 5.976 4.794 5.976 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.050 5.318 arnoldi_extremal 4 6.8 0.000 0.000 4.709 4.725 arnoldi_normal_ev 4 7.8 0.001 0.005 4.709 4.725 build_subspace 16 8.4 0.014 0.021 4.448 4.452 ls_scf_post 1 4.0 0.001 0.006 3.549 3.567 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.182 3.431 ls_scf_store_result 1 5.0 0.000 0.000 3.245 3.369 dbcsr_matrix_vector_mult_local 304 10.0 2.768 3.248 2.770 3.250 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.664 2.778 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.224 2.575 make_images_pack 222 9.7 2.021 2.430 2.023 2.432 mp_allgather_i34 111 8.7 0.987 2.329 0.987 2.329 dbcsr_sort_data 436 11.2 1.822 2.058 1.822 2.058 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.889 1.891 rebuild_ks_matrix 3 7.3 0.000 0.000 1.876 1.878 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.012 1.876 1.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.198000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1748.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 2.786918E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 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 0.047 0.120 95.327 95.328 qs_energies 1 2.0 0.000 0.000 94.506 94.510 ls_scf 1 3.0 0.000 0.000 93.091 93.095 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.151 77.393 ls_scf_main 1 4.0 0.000 0.000 58.000 58.003 multiply_cannon 111 7.7 0.037 0.069 53.118 57.233 multiply_cannon_loop 111 8.7 0.099 0.105 49.542 53.638 density_matrix_trs4 2 5.0 0.002 0.003 51.929 52.146 mp_waitall_1 7281 11.0 24.012 34.009 24.012 34.009 ls_scf_init_scf 1 4.0 0.000 0.001 31.275 31.278 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.022 30.130 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.370 27.382 multiply_cannon_multrec 888 9.7 12.611 15.302 21.390 24.852 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.076 23.125 make_m2s 222 7.7 0.006 0.007 17.218 18.405 make_images 222 8.7 1.974 2.276 17.180 18.367 hybrid_alltoall_any 227 10.6 0.621 2.877 9.472 10.894 make_images_data 222 9.7 0.003 0.004 9.904 10.866 dbcsr_mm_accdrv_process 3754 10.4 0.263 0.593 8.290 9.441 dbcsr_mm_accdrv_process_sort 3754 11.4 7.798 9.007 7.798 9.007 mp_sum_l 887 5.1 5.249 8.780 5.249 8.780 multiply_cannon_sync_h2d 888 9.7 5.995 7.208 5.995 7.208 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.462 7.023 mp_irecv_dv 2335 11.1 2.445 6.976 2.445 6.976 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.030 6.968 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.653 6.862 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.219 5.399 arnoldi_extremal 4 6.8 0.000 0.000 5.158 5.175 arnoldi_normal_ev 4 7.8 0.001 0.004 5.158 5.175 build_subspace 16 8.4 0.014 0.020 4.857 4.863 calculate_norms 1584 9.8 4.307 4.710 4.307 4.710 mp_allgather_i34 111 8.7 1.416 3.893 1.416 3.893 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.486 3.832 ls_scf_post 1 4.0 0.000 0.000 3.815 3.820 dbcsr_matrix_vector_mult_local 304 10.0 3.041 3.625 3.043 3.627 ls_scf_store_result 1 5.0 0.000 0.000 3.257 3.345 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.895 3.006 dbcsr_sort_data 325 11.1 1.912 2.157 1.912 2.157 make_images_sizes 222 9.7 0.000 0.000 0.970 2.121 mp_alltoall_i44 222 10.7 0.970 2.120 0.970 2.120 make_images_pack 222 9.7 1.811 2.081 1.813 2.084 dbcsr_data_release 9322 10.9 1.336 2.012 1.336 2.012 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.919 1.921 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.328000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2164.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 3.394355E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 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 0.036 0.050 98.266 98.267 qs_energies 1 2.0 0.000 0.000 97.575 97.672 ls_scf 1 3.0 0.000 0.000 95.911 96.010 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.928 79.203 ls_scf_main 1 4.0 0.000 0.000 59.232 59.235 multiply_cannon 111 7.7 0.060 0.215 51.899 56.598 density_matrix_trs4 2 5.0 0.002 0.003 53.048 53.177 multiply_cannon_loop 111 8.7 0.115 0.125 46.848 50.464 ls_scf_init_scf 1 4.0 0.001 0.005 33.138 33.139 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.640 31.705 mp_waitall_1 6369 11.0 22.996 29.695 22.996 29.695 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.178 29.188 multiply_cannon_multrec 1332 9.7 14.163 17.295 22.120 24.839 make_m2s 222 7.7 0.007 0.008 21.197 22.638 make_images 222 8.7 3.139 3.602 21.146 22.589 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.405 17.575 make_images_data 222 9.7 0.004 0.004 11.850 13.524 hybrid_alltoall_any 227 10.6 0.797 3.769 11.177 12.977 dbcsr_mm_accdrv_process 3641 10.4 0.215 0.408 7.598 9.139 dbcsr_mm_accdrv_process_sort 3641 11.4 7.216 8.723 7.216 8.723 mp_sum_l 887 5.1 4.250 7.640 4.250 7.640 multiply_cannon_sync_h2d 1332 9.7 5.514 6.232 5.514 6.232 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.572 6.134 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.295 6.036 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.086 6.023 mp_irecv_dv 3229 10.9 2.063 5.984 2.063 5.984 arnoldi_extremal 4 6.8 0.000 0.000 5.336 5.348 arnoldi_normal_ev 4 7.8 0.001 0.005 5.336 5.348 build_subspace 16 8.4 0.014 0.021 4.996 5.004 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.574 4.771 calculate_norms 2376 9.8 4.186 4.539 4.186 4.539 mp_allgather_i34 111 8.7 2.173 4.461 2.173 4.461 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.647 3.909 dbcsr_matrix_vector_mult_local 304 10.0 3.216 3.710 3.218 3.712 ls_scf_post 1 4.0 0.000 0.000 3.541 3.638 dbcsr_sort_data 658 11.4 3.021 3.304 3.021 3.304 dbcsr_special_finalize 555 9.7 0.006 0.007 2.762 3.143 dbcsr_merge_single_wm 555 10.7 0.531 0.649 2.753 3.134 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.053 3.133 ls_scf_store_result 1 5.0 0.000 0.000 2.997 3.069 dbcsr_data_release 10477 10.7 1.592 2.469 1.592 2.469 dbcsr_finalize 304 7.8 0.049 0.061 1.801 1.968 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.267000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2750.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 4.723081E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 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 0.042 0.057 100.249 100.250 qs_energies 1 2.0 0.000 0.000 99.483 99.487 ls_scf 1 3.0 0.000 0.000 97.324 97.333 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.605 78.828 ls_scf_main 1 4.0 0.000 0.000 62.537 62.537 multiply_cannon 111 7.7 0.089 0.245 55.927 61.351 density_matrix_trs4 2 5.0 0.002 0.002 55.337 55.427 multiply_cannon_loop 111 8.7 0.070 0.087 51.347 52.878 mp_waitall_1 5436 11.0 26.856 32.769 26.856 32.769 ls_scf_init_scf 1 4.0 0.000 0.001 31.160 31.164 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.866 29.898 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.672 27.683 multiply_cannon_multrec 444 9.7 14.014 16.252 21.150 24.356 make_m2s 222 7.7 0.004 0.005 17.853 20.389 make_images 222 8.7 3.719 4.416 17.791 20.329 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.482 16.485 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.037 13.591 make_images_data 222 9.7 0.003 0.004 10.133 12.511 hybrid_alltoall_any 227 10.6 0.787 3.777 9.798 12.329 multiply_cannon_sync_h2d 444 9.7 6.476 8.685 6.476 8.685 dbcsr_mm_accdrv_process 3003 10.4 0.187 0.346 6.832 7.965 dbcsr_mm_accdrv_process_sort 3003 11.4 6.518 7.620 6.518 7.620 mp_allgather_i34 111 8.7 2.804 7.099 2.804 7.099 arnoldi_extremal 4 6.8 0.000 0.000 5.957 5.967 arnoldi_normal_ev 4 7.8 0.001 0.004 5.957 5.967 build_subspace 16 8.4 0.015 0.020 5.572 5.581 mp_sum_l 887 5.1 2.882 4.842 2.882 4.842 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.640 4.786 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.264 4.496 dbcsr_matrix_vector_mult_local 304 10.0 3.741 4.212 3.743 4.214 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.617 4.037 mp_irecv_dv 1241 11.2 1.597 4.018 1.597 4.018 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.599 3.676 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.005 3.672 calculate_norms 792 9.8 3.553 3.672 3.553 3.672 ls_scf_post 1 4.0 0.000 0.000 3.628 3.631 make_images_sizes 222 9.7 0.000 0.000 1.070 3.608 mp_alltoall_i44 222 10.7 1.070 3.608 1.070 3.608 ls_scf_store_result 1 5.0 0.000 0.000 3.406 3.452 dbcsr_finalize 304 7.8 0.062 0.077 2.203 2.289 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.144 2.144 dbcsr_merge_all 275 8.9 0.474 0.527 2.051 2.131 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.099 2.100 rebuild_ks_matrix 3 7.3 0.000 0.000 2.065 2.066 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.002 2.065 2.066 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.250000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3678.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 8.867222E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 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 0.111 0.138 109.842 109.843 qs_energies 1 2.0 0.000 0.000 108.106 108.120 ls_scf 1 3.0 0.000 0.000 105.196 105.209 dbcsr_multiply_generic 111 6.7 0.023 0.028 78.135 78.254 ls_scf_main 1 4.0 0.000 0.000 66.475 66.476 density_matrix_trs4 2 5.0 0.002 0.003 57.217 57.288 multiply_cannon 111 7.7 0.121 0.211 50.264 52.002 multiply_cannon_loop 111 8.7 0.068 0.071 46.873 47.865 ls_scf_init_scf 1 4.0 0.001 0.001 34.626 34.627 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.987 32.994 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.126 30.138 mp_waitall_1 4527 11.1 22.679 26.397 22.679 26.397 make_m2s 222 7.7 0.005 0.005 23.993 24.920 make_images 222 8.7 4.590 4.978 23.887 24.812 multiply_cannon_multrec 444 9.7 17.828 18.751 22.503 23.210 hybrid_alltoall_any 227 10.6 1.665 3.643 12.918 15.863 make_images_data 222 9.7 0.003 0.003 13.208 15.626 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.876 11.597 multiply_cannon_sync_h2d 444 9.7 8.846 8.887 8.846 8.887 arnoldi_extremal 4 6.8 0.000 0.000 7.538 7.553 arnoldi_normal_ev 4 7.8 0.002 0.008 7.538 7.553 build_subspace 16 8.4 0.026 0.036 6.997 7.010 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.575 5.736 dbcsr_matrix_vector_mult_local 304 10.0 5.115 5.420 5.117 5.423 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.069 5.302 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.200 5.291 dbcsr_mm_accdrv_process 1814 10.4 0.210 0.317 4.503 4.644 dbcsr_mm_accdrv_process_sort 1814 11.4 4.193 4.327 4.193 4.327 ls_scf_post 1 4.0 0.000 0.000 4.095 4.107 make_images_sizes 222 9.7 0.000 0.000 1.455 3.650 mp_alltoall_i44 222 10.7 1.454 3.650 1.454 3.650 ls_scf_store_result 1 5.0 0.000 0.000 3.462 3.472 mp_allgather_i34 111 8.7 1.036 3.302 1.036 3.302 calculate_norms 792 9.8 3.242 3.282 3.242 3.282 dbcsr_finalize 304 7.8 0.082 0.090 3.104 3.170 dbcsr_merge_all 275 8.9 0.889 0.921 2.875 2.948 dbcsr_complete_redistribute 5 7.6 1.427 1.469 2.776 2.884 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.880 2.881 dbcsr_data_release 12724 10.6 2.340 2.877 2.340 2.877 matrix_ls_to_qs 2 6.0 0.000 0.000 2.412 2.531 dbcsr_sort_data 325 11.1 2.445 2.498 2.445 2.498 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.487 2.489 rebuild_ks_matrix 3 7.3 0.000 0.000 2.421 2.422 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.421 2.422 mp_sum_l 887 5.1 1.552 2.334 1.552 2.334 dbcsr_new_transposed 4 7.5 0.243 0.252 2.271 2.281 dbcsr_frobenius_norm 74 6.6 2.057 2.138 2.189 2.230 dbcsr_add_d 103 6.2 0.000 0.000 2.161 2.229 dbcsr_add_anytype 103 7.2 0.859 0.891 2.161 2.228 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.843000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6835.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb_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 585.773056E+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 76559. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 3741931. 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 0.048 0.091 299.716 299.717 qs_mol_dyn_low 1 2.0 0.003 0.020 298.144 298.546 qs_forces 5 3.8 0.013 0.318 297.781 297.783 qs_energies 5 4.8 0.013 0.401 293.779 294.052 scf_env_do_scf 5 5.8 0.001 0.022 272.913 272.916 scf_env_do_scf_inner_loop 105 6.6 0.003 0.044 235.211 235.224 qs_scf_new_mos 105 7.6 0.000 0.001 180.150 180.330 qs_scf_loop_do_ot 105 8.6 0.001 0.001 180.149 180.329 ot_scf_mini 105 9.6 0.003 0.005 170.143 170.324 dbcsr_multiply_generic 1445 12.2 0.128 0.136 140.941 141.531 velocity_verlet 4 3.0 0.002 0.038 123.505 123.506 multiply_cannon 1445 13.2 0.275 0.287 118.771 121.602 multiply_cannon_loop 1445 14.2 2.377 2.509 116.053 117.630 qs_ot_get_p 112 10.4 0.001 0.001 76.785 77.081 qs_ot_p2m_diag 40 11.0 0.020 0.030 64.137 64.210 ot_mini 105 10.6 0.001 0.001 62.742 62.878 cp_dbcsr_syevd 40 12.0 0.002 0.002 60.578 60.582 cp_fm_syevd 40 13.0 0.000 0.001 54.106 54.274 cp_fm_redistribute_end 40 14.0 24.181 48.216 24.193 48.223 cp_fm_syevd_base 40 14.0 24.006 48.059 24.006 48.059 mp_waitall_1 488190 16.1 39.174 46.160 39.174 46.160 multiply_cannon_multrec 69360 15.2 31.651 37.078 39.416 45.075 qs_ot_get_derivative 55 11.6 0.001 0.001 40.952 41.077 init_scf_loop 7 6.6 0.001 0.062 37.653 37.657 rebuild_ks_matrix 110 8.4 0.000 0.000 35.149 35.309 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.055 35.148 35.308 multiply_cannon_sync_h2d 69360 15.2 30.587 35.087 30.587 35.087 multiply_cannon_metrocomm3 69360 15.2 0.195 0.205 25.377 33.213 qs_ks_update_qs_env 112 7.6 0.001 0.001 32.349 32.488 prepare_preconditioner 7 7.6 0.000 0.001 31.582 31.620 make_preconditioner 7 8.6 0.001 0.041 31.582 31.620 qs_rho_update_rho_low 110 7.6 0.001 0.012 27.660 28.016 calculate_rho_elec 110 8.6 0.030 0.041 27.659 28.016 rs_pw_transfer 690 11.5 0.012 0.041 25.024 26.300 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.136 23.392 apply_single 62 13.6 0.000 0.000 23.136 23.392 make_full_inverse_cholesky 7 9.6 0.000 0.000 22.836 22.914 density_rs2pw 110 9.6 0.006 0.007 21.137 22.353 ot_new_cg_direction 55 11.6 0.001 0.001 21.094 21.096 init_scf_run 5 5.8 0.000 0.001 17.828 17.829 scf_env_initial_rho_setup 5 6.8 0.000 0.002 17.828 17.829 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.169 15.318 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.754 14.872 cp_fm_cholesky_invert 7 10.6 14.772 14.789 14.772 14.789 pw_transfer 1645 12.4 0.082 0.111 13.828 14.136 fft_wrap_pw1pw2 1425 13.5 0.013 0.017 13.686 13.996 sum_up_and_integrate 60 10.3 0.028 0.050 13.670 13.698 integrate_v_rspace 60 11.3 0.002 0.018 13.641 13.670 mp_sum_l 4764 12.2 12.544 13.417 12.544 13.417 calculate_first_density_matrix 1 7.0 0.000 0.007 12.030 12.054 calculate_dm_sparse 110 9.5 0.000 0.001 11.827 12.022 fft_wrap_pw1pw2_240 915 15.0 1.179 1.268 11.389 11.680 check_diag 80 13.5 8.587 8.863 11.404 11.562 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 11.478 11.496 acc_transpose_blocks 69360 15.2 0.352 0.379 10.941 11.323 qs_vxc_create 110 10.4 0.002 0.021 11.242 11.296 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 10.934 10.991 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.905 10.974 rs_pw_transfer_RS2PW_30 110 11.6 1.581 1.659 9.794 10.962 mp_sendrecv_dv 168740 12.6 10.417 10.615 10.417 10.615 make_m2s 2890 13.2 0.078 0.085 9.214 9.824 fft3d_pb 915 16.0 2.365 2.635 9.482 9.797 make_images 2890 14.2 0.238 0.258 9.107 9.717 potential_pw2rs 60 12.3 0.003 0.003 9.517 9.633 acc_transpose_blocks_kernels 69360 16.2 0.844 0.907 9.043 9.286 dbcsr_mm_accdrv_process 154766 15.8 4.013 4.255 7.634 9.168 make_full_single_inverse 7 9.6 0.001 0.002 8.386 8.423 jit_kernel_transpose 5 15.0 8.198 8.394 8.198 8.394 xc_rho_set_and_dset_create 110 12.4 0.078 0.151 7.791 8.049 multiply_cannon_metrocomm1 69360 15.2 0.089 0.096 4.504 7.801 xc_vxc_pw_create 60 11.3 0.040 0.099 7.636 7.690 cp_fm_cholesky_decompose 14 10.2 7.443 7.452 7.443 7.452 mp_alltoall_z22v 2340 17.7 6.650 7.060 6.650 7.060 xc_pw_derive 510 13.4 0.005 0.007 6.938 7.007 mp_waitany 7680 13.5 4.771 6.065 4.771 6.065 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=299.717000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=557.800000, yerr=1.600000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: d0ac4d4144bfa58d4eaeb554dbb276aba7b8aebb Summary: empty Status: OK