=== 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: b57c5e43b9bf4ebed0002f204f6eebfbc3082088 ################# 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 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # 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 (03.02.2023) # # \ 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 ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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_prefixed_scalapack.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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/01 job id: 45981196 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/02 job id: 45981197 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/03 job id: 45981198 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/04 job id: 45981199 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/05 job id: 45981200 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/06 job id: 45981201 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/07 job id: 45981202 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/08 job id: 45981203 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/09 job id: 45981204 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/10 job id: 45981205 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/11 job id: 45981206 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/12 job id: 45981208 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/13 job id: 45981209 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/14 job id: 45981211 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/15 job id: 45981212 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/16 job id: 45981213 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/17 job id: 45981214 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/18 job id: 45981215 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/19 job id: 45981216 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/20 job id: 45981217 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/21 job id: 45981218 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/22 job id: 45981219 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/23 job id: 45981220 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/24 job id: 45981221 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/25 job id: 45981222 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/26 job id: 45981223 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/27 job id: 45981224 --- 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/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 344 9. 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 0.034 134.094 134.095 farming_run 1 2.0 133.635 133.639 134.065 134.068 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.449751E+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 491 2254389. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.026 115.545 115.546 qs_energies 1 2.0 0.000 0.000 115.349 115.351 mp2_main 1 3.0 0.000 0.000 113.139 113.142 mp2_gpw_main 1 4.0 0.019 0.026 112.092 112.095 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 93.135 93.377 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.389 55.632 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.171 41.688 46.621 get_2c_integrals 1 6.0 0.000 0.000 37.283 37.574 integrate_v_rspace 273 8.0 0.440 0.450 25.083 29.708 pw_transfer 6555 10.6 0.374 0.384 27.349 28.124 fft_wrap_pw1pw2 5465 11.4 0.046 0.049 25.999 26.610 grid_integrate_task_list 273 9.0 20.907 25.952 20.907 25.952 fft_wrap_pw1pw2_100 2178 12.4 1.198 1.469 23.525 24.166 compute_2c_integrals 1 7.0 0.002 0.003 19.826 19.827 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.887 19.552 mp2_eri_2c_integrate_gpw 1 9.0 2.385 2.437 18.884 19.548 rpa_ri_compute_en 1 5.0 0.001 0.001 18.850 18.953 cp_fm_cholesky_decompose 12 8.2 17.396 17.711 17.396 17.711 cholesky_decomp 1 7.0 0.000 0.000 16.298 16.605 fft3d_s 5443 13.4 16.132 16.383 16.154 16.404 ao_to_mo_and_store_B_mult_1 272 7.0 10.866 15.605 10.866 15.605 calculate_wavefunction 272 8.0 5.438 5.521 12.528 13.216 rpa_num_int 1 6.0 0.000 0.001 10.525 10.526 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.494 10.524 calc_mat_Q 8 8.0 0.000 0.000 9.358 9.463 contract_S_to_Q 8 9.0 0.000 0.000 8.779 8.881 calc_potential_gpw 544 9.5 0.005 0.005 8.225 8.662 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.203 8.616 parallel_gemm_fm 14 9.1 0.000 0.000 8.361 8.450 parallel_gemm_fm_cosma 14 10.1 8.361 8.450 8.361 8.450 potential_pw2rs 545 10.0 0.107 0.109 7.688 8.363 create_integ_mat 1 6.0 0.022 0.028 7.986 7.987 collocate_single_gaussian 272 10.0 0.039 0.041 7.478 7.846 array2fm 1 7.0 0.000 0.000 6.786 7.433 pw_scatter_s 2720 13.7 4.408 4.604 4.408 4.604 pw_gather_s 2722 13.2 3.856 4.169 3.856 4.169 array2fm_buffer_send 1 8.0 2.977 3.157 2.977 3.157 pw_poisson_solve 545 10.5 1.104 1.152 2.159 2.312 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.091605, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2721.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 344 10. 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 0.029 0.036 397.654 397.656 farming_run 1 2.0 396.998 397.004 397.619 397.621 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.228448E+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 1941 22272. MP_Sync 37 MP_Alltoall 77 12066451. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.030 210.068 210.069 qs_energies 1 2.0 0.000 0.000 209.822 209.829 scf_env_do_scf 1 3.0 0.000 0.000 106.897 106.898 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.893 105.904 rebuild_ks_matrix 4 6.0 0.000 0.000 105.892 105.902 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.068 105.892 105.902 hfx_ks_matrix 4 8.0 0.001 0.001 105.497 105.502 integrate_four_center 4 9.0 0.143 0.468 105.496 105.501 mp2_main 1 3.0 0.000 0.000 102.628 102.634 mp2_gpw_main 1 4.0 0.032 0.046 101.640 101.648 integrate_four_center_main 4 10.0 0.140 0.661 96.980 100.105 integrate_four_center_bin 264 11.0 96.840 100.103 96.840 100.103 init_scf_loop 1 4.0 0.000 0.000 92.588 92.589 mp2_ri_gpw_compute_in 1 5.0 0.063 0.064 74.841 76.084 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.574 55.817 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.162 42.351 47.363 integrate_v_rspace 95 8.0 0.398 0.563 28.733 33.735 pw_transfer 2240 10.6 0.145 0.172 29.879 30.256 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.913 29.254 ao_to_mo_and_store_B_mult_1 91 7.0 10.535 29.216 10.535 29.216 grid_integrate_task_list 95 9.0 24.018 29.186 24.018 29.186 mp2_ri_gpw_compute_en 1 5.0 0.055 0.062 26.653 28.473 fft_wrap_pw1pw2_100 730 12.4 1.236 1.387 26.617 26.961 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.845 1.893 24.768 24.778 get_2c_integrals 1 6.0 0.000 0.000 20.175 20.203 compute_2c_integrals 1 7.0 0.002 0.003 19.156 19.159 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.752 19.033 mp2_eri_2c_integrate_gpw 1 9.0 1.749 1.831 18.751 19.031 fft3d_s 1823 13.4 18.428 18.767 18.441 18.780 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.305 14.305 calculate_wavefunction 91 8.0 2.013 2.050 9.744 9.990 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.589 8.722 9.124 potential_pw2rs 186 10.0 0.033 0.035 8.617 9.103 local_gemm 172 8.0 8.167 8.557 8.167 8.557 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.000 0.001 8.167 8.503 mp2_ri_gpw_compute_en_comm 22 7.0 0.505 0.522 7.805 8.391 calc_potential_gpw 182 9.5 0.002 0.002 7.862 8.096 collocate_single_gaussian 91 10.0 0.017 0.020 7.838 8.061 mp2_ri_gpw_compute_en_ener 172 7.0 6.350 6.431 6.350 6.431 mp_sendrecv_dm3 2068 8.0 5.827 6.405 5.827 6.405 mp_sync 37 10.5 3.174 6.210 3.174 6.210 pw_gather_s 912 13.2 4.928 5.289 4.928 5.289 pw_scatter_s 910 13.7 3.922 4.280 3.922 4.280 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.629343, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1515.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 451.706880E+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 10249 272. 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 0.010 0.029 53.727 53.728 qs_mol_dyn_low 1 2.0 0.003 0.003 53.512 53.520 qs_forces 11 3.9 0.002 0.003 53.445 53.447 qs_energies 11 4.9 0.001 0.001 51.953 51.964 scf_env_do_scf 11 5.9 0.000 0.001 46.023 46.023 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 43.899 43.900 qs_scf_new_mos 108 7.5 0.000 0.001 33.542 33.842 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.542 33.842 dbcsr_multiply_generic 2286 12.5 0.094 0.097 33.006 33.374 ot_scf_mini 108 9.5 0.002 0.002 31.842 32.076 multiply_cannon 2286 13.5 0.182 0.192 25.697 27.159 multiply_cannon_loop 2286 14.5 1.490 1.573 24.967 26.503 velocity_verlet 10 3.0 0.001 0.001 26.129 26.129 ot_mini 108 10.5 0.001 0.001 18.882 19.160 qs_ot_get_derivative 108 11.5 0.001 0.001 16.017 16.223 mp_waitall_1 245248 16.5 8.380 14.605 8.380 14.605 multiply_cannon_metrocomm3 54864 15.5 0.069 0.075 5.857 12.597 multiply_cannon_multrec 54864 15.5 4.247 6.606 7.485 10.878 qs_ot_get_p 119 10.4 0.001 0.001 8.326 8.585 rebuild_ks_matrix 119 8.3 0.000 0.000 8.026 8.181 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.025 8.181 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.072 7.212 multiply_cannon_sync_h2d 54864 15.5 5.927 7.029 5.927 7.029 mp_sum_l 7207 12.9 5.240 7.024 5.240 7.024 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.527 5.989 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.489 5.535 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.318 5.415 init_scf_run 11 5.9 0.000 0.001 4.675 4.675 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.675 4.675 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.639 4.639 sum_up_and_integrate 119 10.3 0.012 0.014 4.624 4.631 integrate_v_rspace 119 11.3 0.002 0.003 4.612 4.621 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.356 4.484 calculate_rho_elec 119 8.7 0.012 0.017 4.355 4.484 dbcsr_mm_accdrv_process 76910 16.1 1.155 1.835 3.160 4.444 cp_fm_diag_elpa 50 13.0 0.000 0.001 4.391 4.392 cp_fm_redistribute_end 50 14.0 2.239 4.360 2.249 4.363 cp_fm_diag_elpa_base 50 14.0 2.107 4.238 2.112 4.250 calculate_dm_sparse 119 9.5 0.000 0.001 2.938 3.070 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.888 3.066 apply_single 119 13.6 0.000 0.000 2.888 3.066 rs_pw_transfer 974 11.9 0.011 0.012 2.893 2.992 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.763 2.863 ot_diis_step 108 11.5 0.006 0.006 2.636 2.636 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.437 2.540 jit_kernel_multiply 13 15.8 1.945 2.524 1.945 2.524 density_rs2pw 119 9.7 0.004 0.004 2.361 2.475 calculate_first_density_matrix 1 7.0 0.000 0.000 2.454 2.458 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.248 2.251 acc_transpose_blocks 54864 15.5 0.229 0.256 1.731 2.199 wfi_extrapolate 11 7.9 0.001 0.001 2.157 2.157 grid_integrate_task_list 119 12.3 2.031 2.117 2.031 2.117 init_scf_loop 11 6.9 0.000 0.000 2.105 2.106 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.947 2.006 potential_pw2rs 119 12.3 0.004 0.004 1.957 1.971 pw_transfer 1439 11.6 0.052 0.055 1.844 1.919 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.769 1.847 make_m2s 4572 13.5 0.054 0.056 1.753 1.791 mp_sum_d 4135 12.0 1.214 1.773 1.214 1.773 make_images 4572 14.5 0.133 0.138 1.671 1.708 mp_alltoall_d11v 2130 13.8 1.325 1.501 1.325 1.501 fft3d_ps 1201 14.6 0.375 0.482 1.413 1.493 fft_wrap_pw1pw2_140 487 13.2 0.183 0.201 1.372 1.453 mp_waitany 12084 13.8 1.236 1.434 1.236 1.434 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.406 1.424 grid_collocate_task_list 119 9.7 1.296 1.383 1.296 1.383 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.728000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.454545, yerr=0.890724 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.194048E+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 10226 305. MP_Sync 54 MP_Alltoall 2060 1032702. 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 0.010 0.028 39.633 39.633 qs_mol_dyn_low 1 2.0 0.003 0.003 39.418 39.440 qs_forces 11 3.9 0.002 0.002 39.356 39.357 qs_energies 11 4.9 0.001 0.001 37.660 37.663 scf_env_do_scf 11 5.9 0.000 0.001 32.485 32.486 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.930 29.930 dbcsr_multiply_generic 2286 12.5 0.103 0.108 21.437 21.798 qs_scf_new_mos 108 7.5 0.001 0.001 20.447 20.682 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.446 20.681 ot_scf_mini 108 9.5 0.002 0.003 19.548 19.724 velocity_verlet 10 3.0 0.001 0.001 18.569 18.570 multiply_cannon 2286 13.5 0.205 0.215 16.327 17.890 multiply_cannon_loop 2286 14.5 0.906 0.978 15.200 16.536 ot_mini 108 10.5 0.001 0.001 11.867 12.102 mp_waitall_1 200699 16.5 5.746 10.988 5.746 10.988 qs_ot_get_derivative 108 11.5 0.001 0.001 9.402 9.580 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 4.111 9.305 multiply_cannon_multrec 27432 15.5 1.986 4.423 5.894 8.622 rebuild_ks_matrix 119 8.3 0.000 0.000 7.371 7.508 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.370 7.508 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.525 6.650 dbcsr_mm_accdrv_process 47894 16.0 3.044 4.936 3.839 5.663 qs_ot_get_p 119 10.4 0.001 0.001 4.783 5.014 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.601 4.455 sum_up_and_integrate 119 10.3 0.024 0.027 4.347 4.354 integrate_v_rspace 119 11.3 0.002 0.003 4.323 4.331 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.050 4.128 apply_single 119 13.6 0.000 0.000 3.049 4.128 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.979 4.010 calculate_rho_elec 119 8.7 0.021 0.024 3.979 4.010 mp_sum_l 7207 12.9 2.055 3.993 2.055 3.993 init_scf_run 11 5.9 0.000 0.001 3.960 3.961 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.960 3.961 rs_pw_transfer 974 11.9 0.009 0.011 2.827 3.294 qs_ot_p2m_diag 50 11.0 0.009 0.012 3.250 3.268 make_m2s 4572 13.5 0.052 0.054 2.599 2.818 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.809 2.810 multiply_cannon_sync_h2d 27432 15.5 2.186 2.796 2.186 2.796 density_rs2pw 119 9.7 0.004 0.004 2.294 2.764 make_images 4572 14.5 0.200 0.236 2.510 2.726 init_scf_loop 11 6.9 0.000 0.000 2.532 2.532 cp_fm_diag_elpa 50 13.0 0.000 0.001 2.436 2.436 ot_diis_step 108 11.5 0.011 0.011 2.416 2.417 cp_fm_redistribute_end 50 14.0 1.231 2.396 1.237 2.400 calculate_first_density_matrix 1 7.0 0.000 0.000 2.376 2.390 cp_fm_diag_elpa_base 50 14.0 1.129 2.290 1.158 2.333 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.202 2.292 calculate_dm_sparse 119 9.5 0.000 0.001 2.114 2.202 potential_pw2rs 119 12.3 0.006 0.006 2.066 2.080 pw_transfer 1439 11.6 0.065 0.069 2.022 2.055 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.930 1.964 grid_integrate_task_list 119 12.3 1.837 1.932 1.837 1.932 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.885 1.888 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.656 1.696 prepare_preconditioner 11 7.9 0.000 0.000 1.582 1.608 make_preconditioner 11 8.9 0.000 0.000 1.582 1.608 make_images_data 4572 15.5 0.045 0.051 1.170 1.602 jit_kernel_multiply 8 16.2 0.743 1.595 0.743 1.595 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.469 1.527 acc_transpose_blocks 27432 15.5 0.109 0.114 1.220 1.526 fft_wrap_pw1pw2_140 487 13.2 0.202 0.212 1.481 1.517 wfi_extrapolate 11 7.9 0.001 0.001 1.515 1.515 hybrid_alltoall_any 4725 16.4 0.051 0.113 1.015 1.513 fft3d_ps 1201 14.6 0.523 0.583 1.481 1.510 grid_collocate_task_list 119 9.7 1.227 1.371 1.227 1.371 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.328 1.347 mp_alltoall_d11v 2130 13.8 1.241 1.337 1.241 1.337 mp_allgather_i34 2286 14.5 0.571 1.317 0.571 1.317 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.178 1.226 mp_sum_d 4135 12.0 0.602 1.043 0.602 1.043 mp_waitany 5720 13.7 0.523 1.022 0.523 1.022 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.147 0.542 0.997 qs_energies_init_hamiltonians 11 5.9 0.003 0.008 0.947 0.948 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.904 0.918 acc_transpose_blocks_kernels 27432 16.5 0.182 0.270 0.678 0.901 make_images_sizes 4572 15.5 0.005 0.005 0.613 0.875 mp_alltoall_i44 4572 16.5 0.608 0.870 0.608 0.870 rs_pw_transfer_PW2RS_50 119 14.3 0.586 0.607 0.761 0.811 mp_alltoall_z22v 1201 16.6 0.728 0.808 0.728 0.808 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.633000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.727273, yerr=1.656442 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 521.715712E+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 10075 307. MP_Sync 54 MP_Alltoall 1821 1607886. 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 0.010 0.028 32.570 32.571 qs_mol_dyn_low 1 2.0 0.003 0.003 32.305 32.315 qs_forces 11 3.9 0.002 0.002 32.245 32.246 qs_energies 11 4.9 0.001 0.001 30.672 30.675 scf_env_do_scf 11 5.9 0.000 0.001 25.810 25.810 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 23.288 23.288 dbcsr_multiply_generic 2286 12.5 0.095 0.097 16.694 16.779 velocity_verlet 10 3.0 0.001 0.001 15.246 15.248 qs_scf_new_mos 108 7.5 0.001 0.001 15.065 15.082 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.064 15.082 ot_scf_mini 108 9.5 0.002 0.002 14.346 14.357 multiply_cannon 2286 13.5 0.194 0.200 13.447 14.232 multiply_cannon_loop 2286 14.5 0.637 0.665 12.655 13.476 ot_mini 108 10.5 0.001 0.001 8.840 8.852 qs_ot_get_derivative 108 11.5 0.001 0.001 7.356 7.367 multiply_cannon_multrec 18288 15.5 1.915 2.842 6.955 7.206 rebuild_ks_matrix 119 8.3 0.000 0.000 6.390 6.409 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.390 6.409 dbcsr_mm_accdrv_process 38222 16.0 4.091 5.649 4.955 5.855 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.646 5.663 sum_up_and_integrate 119 10.3 0.031 0.032 3.957 3.963 integrate_v_rspace 119 11.3 0.003 0.003 3.926 3.936 mp_waitall_1 158411 16.6 2.727 3.903 2.727 3.903 init_scf_run 11 5.9 0.000 0.001 3.662 3.662 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.662 3.662 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.484 3.491 calculate_rho_elec 119 8.7 0.031 0.031 3.483 3.491 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.877 3.479 qs_ot_get_p 119 10.4 0.001 0.001 3.339 3.368 rs_pw_transfer 974 11.9 0.009 0.010 2.339 2.602 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.491 2.527 init_scf_loop 11 6.9 0.000 0.000 2.505 2.506 calculate_first_density_matrix 1 7.0 0.000 0.000 2.437 2.438 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.015 2.338 apply_single 119 13.6 0.000 0.000 2.015 2.338 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.260 2.265 density_rs2pw 119 9.7 0.004 0.004 1.996 2.259 jit_kernel_multiply 10 16.0 0.813 2.076 0.813 2.076 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.987 1.987 make_m2s 4572 13.5 0.044 0.045 1.799 1.951 calculate_dm_sparse 119 9.5 0.000 0.001 1.896 1.909 pw_transfer 1439 11.6 0.065 0.069 1.895 1.906 grid_integrate_task_list 119 12.3 1.812 1.891 1.812 1.891 make_images 4572 14.5 0.189 0.201 1.714 1.864 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.802 1.815 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.813 1.814 cp_fm_diag_elpa 50 13.0 0.001 0.003 1.728 1.740 cp_fm_diag_elpa_base 50 14.0 1.701 1.716 1.723 1.736 potential_pw2rs 119 12.3 0.007 0.008 1.712 1.717 prepare_preconditioner 11 7.9 0.000 0.000 1.707 1.709 make_preconditioner 11 8.9 0.000 0.000 1.707 1.709 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.682 1.687 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.567 1.648 multiply_cannon_sync_h2d 18288 15.5 1.425 1.604 1.425 1.604 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.576 1.582 ot_diis_step 108 11.5 0.011 0.011 1.466 1.466 fft_wrap_pw1pw2_140 487 13.2 0.252 0.259 1.453 1.465 mp_sum_l 7207 12.9 1.059 1.406 1.059 1.406 acc_transpose_blocks 18288 15.5 0.076 0.078 1.313 1.376 grid_collocate_task_list 119 9.7 1.214 1.365 1.214 1.365 fft3d_ps 1201 14.6 0.536 0.555 1.291 1.306 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.269 1.273 wfi_extrapolate 11 7.9 0.001 0.001 1.177 1.177 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.990 0.991 make_images_data 4572 15.5 0.045 0.049 0.815 0.988 acc_transpose_blocks_kernels 18288 16.5 0.211 0.220 0.865 0.922 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.697 0.888 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.867 0.888 mp_waitany 9880 13.7 0.568 0.853 0.568 0.853 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.121 0.556 0.821 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.814 0.816 mp_alltoall_d11v 2130 13.8 0.659 0.777 0.659 0.777 cp_fm_cholesky_invert 11 10.9 0.725 0.728 0.725 0.728 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.660 0.727 jit_kernel_transpose 5 15.6 0.654 0.702 0.654 0.702 mp_alltoall_z22v 1201 16.6 0.612 0.676 0.612 0.676 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.571000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.818182, yerr=1.402477 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 556.994560E+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 10074 349. 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 0.028 37.453 37.454 qs_mol_dyn_low 1 2.0 0.003 0.003 37.270 37.277 qs_forces 11 3.9 0.002 0.003 37.205 37.209 qs_energies 11 4.9 0.001 0.004 35.467 35.478 scf_env_do_scf 11 5.9 0.000 0.001 28.380 28.382 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 24.906 24.907 dbcsr_multiply_generic 2286 12.5 0.099 0.102 20.056 20.172 velocity_verlet 10 3.0 0.001 0.001 17.955 17.957 multiply_cannon 2286 13.5 0.232 0.276 16.438 16.820 qs_scf_new_mos 108 7.5 0.001 0.001 16.343 16.406 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.343 16.406 multiply_cannon_loop 2286 14.5 0.945 0.974 15.493 15.827 ot_scf_mini 108 9.5 0.002 0.003 15.406 15.457 ot_mini 108 10.5 0.001 0.001 9.238 9.306 multiply_cannon_multrec 27432 15.5 2.337 2.992 8.594 8.973 qs_ot_get_derivative 108 11.5 0.001 0.001 7.432 7.486 dbcsr_mm_accdrv_process 47916 15.9 5.371 7.214 6.163 7.350 rebuild_ks_matrix 119 8.3 0.000 0.000 6.628 6.683 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.627 6.682 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.880 5.930 init_scf_run 11 5.9 0.000 0.002 5.702 5.702 scf_env_initial_rho_setup 11 6.9 0.002 0.005 5.701 5.702 calculate_first_density_matrix 1 7.0 0.001 0.005 4.299 4.302 mp_waitall_1 137007 16.6 3.501 4.298 3.501 4.298 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.862 3.863 sum_up_and_integrate 119 10.3 0.035 0.038 3.850 3.858 integrate_v_rspace 119 11.3 0.003 0.003 3.815 3.824 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.659 3.677 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.643 3.677 calculate_rho_elec 119 8.7 0.040 0.046 3.642 3.676 qs_ot_get_p 119 10.4 0.001 0.001 3.555 3.626 acc_transpose_blocks 27432 15.5 0.114 0.116 1.829 3.526 init_scf_loop 11 6.9 0.001 0.007 3.453 3.455 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.780 3.177 multiply_cannon_metrocomm3 27432 15.5 0.038 0.040 1.519 3.097 acc_transpose_blocks_kernels 27432 16.5 0.269 0.275 1.191 2.879 multiply_cannon_metrocomm1 27432 15.5 0.034 0.035 1.105 2.627 jit_kernel_transpose 5 15.6 0.922 2.606 0.922 2.606 prepare_preconditioner 11 7.9 0.000 0.000 2.581 2.588 make_preconditioner 11 8.9 0.000 0.001 2.581 2.587 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.185 2.515 rs_pw_transfer 974 11.9 0.009 0.009 2.136 2.330 make_m2s 4572 13.5 0.054 0.056 2.221 2.317 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.249 2.258 density_rs2pw 119 9.7 0.004 0.004 2.058 2.243 make_images 4572 14.5 0.271 0.335 2.113 2.206 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.040 2.169 apply_single 119 13.6 0.000 0.000 2.039 2.169 pw_transfer 1439 11.6 0.065 0.070 2.045 2.082 calculate_dm_sparse 119 9.5 0.000 0.000 2.028 2.081 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.952 1.994 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 1.946 1.974 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.920 1.921 grid_integrate_task_list 119 12.3 1.814 1.909 1.814 1.909 ot_diis_step 108 11.5 0.012 0.013 1.764 1.765 cp_fm_diag_elpa 50 13.0 0.001 0.001 1.659 1.669 cp_fm_diag_elpa_base 50 14.0 1.624 1.643 1.657 1.666 jit_kernel_multiply 9 16.1 0.732 1.648 0.732 1.648 fft_wrap_pw1pw2_140 487 13.2 0.287 0.301 1.596 1.639 potential_pw2rs 119 12.3 0.009 0.009 1.592 1.599 fft3d_ps 1201 14.6 0.559 0.608 1.390 1.415 wfi_extrapolate 11 7.9 0.001 0.001 1.350 1.350 grid_collocate_task_list 119 9.7 1.221 1.342 1.221 1.342 mp_sum_l 7207 12.9 0.972 1.265 0.972 1.265 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.217 1.233 cp_fm_upper_to_full 72 14.2 0.821 1.173 0.821 1.173 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.116 1.134 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.104 1.108 dbcsr_complete_redistribute 329 12.2 0.136 0.174 0.809 1.083 multiply_cannon_sync_h2d 27432 15.5 0.980 1.061 0.980 1.061 make_images_data 4572 15.5 0.045 0.049 0.875 0.990 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.747 0.923 mp_alltoall_d11v 2130 13.8 0.754 0.915 0.754 0.915 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.871 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.592 0.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.843 0.847 cp_fm_cholesky_invert 11 10.9 0.766 0.769 0.766 0.769 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.454000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.181818, yerr=3.185712 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 601.714688E+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 10074 348. 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 0.014 0.028 27.886 27.887 qs_mol_dyn_low 1 2.0 0.003 0.003 27.720 27.727 qs_forces 11 3.9 0.003 0.003 27.655 27.656 qs_energies 11 4.9 0.001 0.001 25.945 25.948 scf_env_do_scf 11 5.9 0.000 0.001 21.335 21.335 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 18.828 18.828 velocity_verlet 10 3.0 0.001 0.001 14.381 14.384 dbcsr_multiply_generic 2286 12.5 0.092 0.097 12.041 12.112 qs_scf_new_mos 108 7.5 0.001 0.001 10.848 10.871 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.848 10.870 ot_scf_mini 108 9.5 0.002 0.002 10.182 10.202 multiply_cannon 2286 13.5 0.228 0.232 9.511 9.952 multiply_cannon_loop 2286 14.5 0.330 0.341 8.604 8.804 multiply_cannon_multrec 9144 15.5 1.587 1.818 5.760 6.008 rebuild_ks_matrix 119 8.3 0.000 0.000 5.973 5.992 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 5.973 5.992 ot_mini 108 10.5 0.001 0.001 5.671 5.698 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.325 5.342 qs_ot_get_derivative 108 11.5 0.001 0.001 4.374 4.394 dbcsr_mm_accdrv_process 12550 15.8 2.976 3.754 4.072 4.158 sum_up_and_integrate 119 10.3 0.037 0.041 3.657 3.662 integrate_v_rspace 119 11.3 0.003 0.003 3.620 3.625 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.528 3.537 calculate_rho_elec 119 8.7 0.060 0.061 3.527 3.537 init_scf_run 11 5.9 0.000 0.001 3.186 3.186 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.185 3.185 qs_ot_get_p 119 10.4 0.001 0.001 2.740 2.776 init_scf_loop 11 6.9 0.000 0.000 2.485 2.486 pw_transfer 1439 11.6 0.066 0.069 2.055 2.066 calculate_first_density_matrix 1 7.0 0.000 0.000 2.051 2.053 density_rs2pw 119 9.7 0.004 0.004 1.845 1.986 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.961 1.973 mp_waitall_1 115863 16.7 1.428 1.960 1.428 1.960 grid_integrate_task_list 119 12.3 1.846 1.943 1.846 1.943 make_m2s 4572 13.5 0.035 0.036 1.730 1.885 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.817 1.819 rs_pw_transfer 974 11.9 0.008 0.008 1.668 1.811 make_images 4572 14.5 0.267 0.299 1.640 1.792 jit_kernel_multiply 10 15.8 1.058 1.766 1.058 1.766 prepare_preconditioner 11 7.9 0.000 0.000 1.742 1.746 make_preconditioner 11 8.9 0.000 0.000 1.742 1.746 calculate_dm_sparse 119 9.5 0.000 0.000 1.697 1.715 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.629 1.654 fft_wrap_pw1pw2_140 487 13.2 0.363 0.373 1.611 1.623 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.620 1.620 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.518 1.519 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.504 1.516 grid_collocate_task_list 119 9.7 1.271 1.391 1.271 1.391 potential_pw2rs 119 12.3 0.010 0.011 1.385 1.388 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.367 1.374 cp_fm_diag_elpa 50 13.0 0.001 0.001 1.321 1.330 cp_fm_diag_elpa_base 50 14.0 1.294 1.311 1.319 1.327 fft3d_ps 1201 14.6 0.564 0.577 1.295 1.305 ot_diis_step 108 11.5 0.013 0.013 1.286 1.287 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.221 1.223 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.161 1.180 apply_single 119 13.6 0.000 0.000 1.161 1.180 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.131 1.143 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.123 1.128 wfi_extrapolate 11 7.9 0.001 0.001 1.081 1.081 hybrid_alltoall_any 4725 16.4 0.063 0.177 0.799 1.033 make_images_data 4572 15.5 0.039 0.042 0.814 0.990 mp_alltoall_d11v 2130 13.8 0.792 0.926 0.792 0.926 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.866 0.921 acc_transpose_blocks 9144 15.5 0.038 0.039 0.900 0.907 cp_fm_cholesky_invert 11 10.9 0.867 0.870 0.867 0.870 multiply_cannon_sync_h2d 9144 15.5 0.711 0.785 0.711 0.785 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.770 0.773 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.673 0.725 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.715 0.723 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.363 0.678 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 0.655 0.661 mp_allgather_i34 2286 14.5 0.239 0.660 0.239 0.660 mp_alltoall_z22v 1201 16.6 0.606 0.636 0.606 0.636 mp_sum_l 7207 12.9 0.370 0.610 0.370 0.610 qs_create_task_list 11 7.9 0.000 0.001 0.542 0.571 generate_qs_task_list 11 8.9 0.188 0.210 0.542 0.570 mp_waitany 5200 13.7 0.442 0.564 0.442 0.564 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.887000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=570.727273, yerr=4.574453 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 766.472192E+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 10074 433. 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 0.013 0.034 41.078 41.079 qs_mol_dyn_low 1 2.0 0.003 0.003 40.878 40.885 qs_forces 11 3.9 0.002 0.004 40.810 40.814 qs_energies 11 4.9 0.001 0.001 38.828 38.831 scf_env_do_scf 11 5.9 0.001 0.001 33.124 33.124 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 25.348 25.350 velocity_verlet 10 3.0 0.001 0.001 23.216 23.222 dbcsr_multiply_generic 2286 12.5 0.102 0.105 17.375 17.595 qs_scf_new_mos 108 7.5 0.001 0.001 15.649 15.747 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.648 15.746 ot_scf_mini 108 9.5 0.002 0.002 14.576 14.676 multiply_cannon 2286 13.5 0.301 0.306 13.599 14.566 multiply_cannon_loop 2286 14.5 0.343 0.348 12.319 13.293 ot_mini 108 10.5 0.001 0.001 8.701 8.815 multiply_cannon_multrec 9144 15.5 3.380 4.671 8.596 8.699 init_scf_loop 11 6.9 0.000 0.000 7.748 7.751 rebuild_ks_matrix 119 8.3 0.000 0.000 7.245 7.385 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.245 7.385 prepare_preconditioner 11 7.9 0.000 0.000 6.791 6.804 make_preconditioner 11 8.9 0.000 0.000 6.791 6.804 qs_ot_get_derivative 108 11.5 0.001 0.001 6.681 6.779 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.362 6.683 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.546 6.672 dbcsr_mm_accdrv_process 12550 15.8 4.032 5.563 5.093 6.428 cp_fm_upper_to_full 72 14.2 3.148 4.522 3.148 4.522 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.215 4.220 calculate_rho_elec 119 8.7 0.118 0.121 4.215 4.219 sum_up_and_integrate 119 10.3 0.064 0.066 3.963 3.969 integrate_v_rspace 119 11.3 0.003 0.003 3.898 3.904 init_scf_run 11 5.9 0.000 0.001 3.672 3.673 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.672 3.672 mp_waitall_1 94719 16.7 2.271 3.349 2.271 3.349 qs_ot_get_p 119 10.4 0.001 0.001 3.141 3.270 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.452 2.885 dbcsr_complete_redistribute 329 12.2 0.287 0.293 1.946 2.752 pw_transfer 1439 11.6 0.069 0.069 2.668 2.673 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.570 2.574 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.201 2.470 apply_single 119 13.6 0.000 0.000 2.200 2.470 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.649 2.444 make_m2s 4572 13.5 0.038 0.039 2.237 2.415 make_images 4572 14.5 0.351 0.384 2.115 2.294 calculate_first_density_matrix 1 7.0 0.000 0.000 2.231 2.233 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.260 2.231 mp_alltoall_i22 627 13.8 1.401 2.224 1.401 2.224 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.424 2.211 fft_wrap_pw1pw2_140 487 13.2 0.614 0.617 2.168 2.174 density_rs2pw 119 9.7 0.004 0.004 2.150 2.160 calculate_dm_sparse 119 9.5 0.000 0.000 2.141 2.159 grid_integrate_task_list 119 12.3 2.046 2.060 2.046 2.060 ot_diis_step 108 11.5 0.014 0.014 1.997 1.997 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.923 1.924 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.914 1.915 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.767 1.772 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.676 1.709 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.658 1.708 mp_sum_l 7207 12.9 0.995 1.681 0.995 1.681 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.642 1.643 fft3d_ps 1201 14.6 0.604 0.618 1.590 1.591 jit_kernel_multiply 8 15.4 1.034 1.499 1.034 1.499 grid_collocate_task_list 119 9.7 1.481 1.491 1.481 1.491 rs_pw_transfer 974 11.9 0.009 0.009 1.438 1.464 cp_fm_cholesky_invert 11 10.9 1.456 1.459 1.456 1.459 potential_pw2rs 119 12.3 0.014 0.014 1.378 1.380 wfi_extrapolate 11 7.9 0.001 0.001 1.377 1.377 cp_fm_diag_elpa 50 13.0 0.001 0.001 1.358 1.358 cp_fm_diag_elpa_base 50 14.0 1.212 1.267 1.356 1.356 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.097 1.352 make_images_data 4572 15.5 0.043 0.046 1.053 1.275 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.217 1.224 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.130 1.148 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.100 1.115 mp_alltoall_d11v 2130 13.8 1.049 1.069 1.049 1.069 multiply_cannon_sync_h2d 9144 15.5 1.039 1.042 1.039 1.042 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.985 1.029 acc_transpose_blocks 9144 15.5 0.039 0.039 0.924 0.949 qs_create_task_list 11 7.9 0.000 0.001 0.934 0.947 generate_qs_task_list 11 8.9 0.367 0.387 0.933 0.946 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.911 0.924 mp_alltoall_z22v 1201 16.6 0.850 0.868 0.850 0.868 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.079000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=707.636364, yerr=17.126655 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 505.589760E+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 9696 492. MP_Sync 52 MP_Alltoall 1938 1535858. 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 0.011 0.030 82.652 82.653 qs_mol_dyn_low 1 2.0 0.003 0.003 82.400 82.410 qs_forces 11 3.9 0.003 0.003 82.332 82.333 qs_energies 11 4.9 0.001 0.001 79.442 79.471 scf_env_do_scf 11 5.9 0.000 0.001 68.839 68.841 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 63.348 63.348 dbcsr_multiply_generic 2055 12.4 0.106 0.110 51.858 52.142 qs_scf_new_mos 99 7.5 0.000 0.001 46.010 46.149 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.010 46.149 multiply_cannon 2055 13.4 0.181 0.191 43.842 44.692 multiply_cannon_loop 2055 14.4 1.518 1.563 42.890 43.807 ot_scf_mini 99 9.5 0.002 0.002 43.686 43.770 velocity_verlet 10 3.0 0.001 0.001 42.148 42.149 ot_mini 99 10.5 0.001 0.001 25.785 25.893 qs_ot_get_derivative 99 11.5 0.001 0.001 19.073 19.190 multiply_cannon_multrec 49320 15.4 12.488 13.096 17.366 18.064 rebuild_ks_matrix 110 8.3 0.000 0.000 14.557 14.711 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 14.557 14.711 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.763 12.904 mp_waitall_1 220248 16.4 10.371 11.448 10.371 11.448 multiply_cannon_sync_h2d 49320 15.4 10.364 10.923 10.364 10.923 qs_ot_get_p 110 10.4 0.001 0.001 9.556 9.664 init_scf_run 11 5.9 0.000 0.001 8.587 8.588 scf_env_initial_rho_setup 11 6.9 0.000 0.001 8.587 8.587 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.194 7.682 apply_single 110 13.6 0.000 0.000 7.194 7.682 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.065 7.626 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 6.295 7.315 sum_up_and_integrate 110 10.3 0.036 0.042 7.104 7.117 integrate_v_rspace 110 11.3 0.003 0.004 7.068 7.090 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.572 6.696 calculate_rho_elec 110 8.6 0.021 0.026 6.572 6.696 ot_diis_step 99 11.5 0.006 0.006 6.557 6.557 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.473 6.507 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.645 5.646 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.455 5.510 init_scf_loop 11 6.9 0.000 0.000 5.465 5.465 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.433 5.437 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.102 5.157 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.081 5.100 cp_fm_diag_elpa_base 48 14.0 5.068 5.087 5.080 5.099 dbcsr_mm_accdrv_process 87628 16.1 1.930 2.030 4.752 5.060 calculate_first_density_matrix 1 7.0 0.000 0.000 4.477 4.483 mp_sum_l 6514 12.8 3.734 4.427 3.734 4.427 rs_pw_transfer 902 11.9 0.011 0.013 3.685 4.269 acc_transpose_blocks 49320 15.4 0.210 0.219 3.999 4.075 wfi_extrapolate 11 7.9 0.001 0.001 4.022 4.022 density_rs2pw 110 9.6 0.004 0.005 3.394 4.007 make_m2s 4110 13.4 0.060 0.066 3.817 3.935 make_images 4110 14.4 0.177 0.189 3.721 3.844 calculate_dm_sparse 110 9.5 0.000 0.001 3.686 3.774 multiply_cannon_metrocomm1 49320 15.4 0.062 0.064 2.242 3.510 grid_integrate_task_list 110 12.3 3.238 3.392 3.238 3.392 prepare_preconditioner 11 7.9 0.000 0.000 3.282 3.307 make_preconditioner 11 8.9 0.000 0.000 3.282 3.307 pw_transfer 1331 11.6 0.055 0.068 3.136 3.215 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.098 3.145 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.047 3.129 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.059 3.106 acc_transpose_blocks_kernels 49320 16.4 0.423 0.439 2.967 2.986 fft_wrap_pw1pw2_140 451 13.1 0.454 0.498 2.596 2.682 potential_pw2rs 110 12.3 0.006 0.007 2.629 2.657 jit_kernel_multiply 13 15.9 2.548 2.586 2.548 2.586 jit_kernel_transpose 5 15.6 2.544 2.557 2.544 2.557 mp_waitany 14300 13.8 1.809 2.492 1.809 2.492 mp_alltoall_d11v 2046 13.8 2.049 2.383 2.049 2.383 grid_collocate_task_list 110 9.6 2.085 2.322 2.085 2.322 fft3d_ps 1111 14.6 0.795 0.879 2.241 2.294 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.924 1.942 make_images_data 4110 15.4 0.042 0.045 1.746 1.881 cp_fm_cholesky_invert 11 10.9 1.771 1.775 1.771 1.775 hybrid_alltoall_any 4261 16.3 0.083 0.486 1.520 1.768 mp_sum_d 3889 11.9 1.298 1.757 1.298 1.757 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.640 1.684 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.653000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=478.090909, yerr=3.476010 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 590.929920E+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 9695 570. MP_Sync 52 MP_Alltoall 1717 3653176. 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 0.013 0.035 69.018 69.018 qs_mol_dyn_low 1 2.0 0.003 0.003 68.766 68.776 qs_forces 11 3.9 0.004 0.012 68.614 68.615 qs_energies 11 4.9 0.001 0.004 65.279 65.283 scf_env_do_scf 11 5.9 0.000 0.001 56.700 56.703 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.063 49.063 dbcsr_multiply_generic 2055 12.4 0.115 0.120 37.475 37.597 velocity_verlet 10 3.0 0.001 0.001 36.282 36.285 qs_scf_new_mos 99 7.5 0.001 0.001 32.754 32.894 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.754 32.894 multiply_cannon 2055 13.4 0.220 0.242 30.974 31.910 ot_scf_mini 99 9.5 0.003 0.003 31.101 31.231 multiply_cannon_loop 2055 14.4 0.930 0.955 29.690 30.444 ot_mini 99 10.5 0.001 0.001 18.105 18.251 multiply_cannon_multrec 24660 15.4 7.636 9.640 13.832 15.608 rebuild_ks_matrix 110 8.3 0.000 0.000 13.794 13.905 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.793 13.905 qs_ot_get_derivative 99 11.5 0.001 0.001 12.281 12.414 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.145 12.245 mp_waitall_1 176588 16.5 7.570 10.085 7.570 10.085 multiply_cannon_sync_h2d 24660 15.4 7.070 8.227 7.070 8.227 init_scf_loop 11 6.9 0.001 0.010 7.601 7.601 multiply_cannon_metrocomm3 24660 15.4 0.067 0.071 5.130 7.537 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.547 7.238 apply_single 110 13.6 0.000 0.001 6.546 7.238 sum_up_and_integrate 110 10.3 0.052 0.059 6.627 6.640 integrate_v_rspace 110 11.3 0.002 0.003 6.575 6.588 qs_ot_get_p 110 10.4 0.001 0.001 6.363 6.535 dbcsr_mm_accdrv_process 52282 16.1 4.661 5.566 6.034 6.305 init_scf_run 11 5.9 0.000 0.002 6.186 6.186 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.185 6.186 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.173 6.184 calculate_rho_elec 110 8.6 0.040 0.048 6.173 6.183 ot_diis_step 99 11.5 0.010 0.011 5.778 5.778 prepare_preconditioner 11 7.9 0.000 0.000 5.567 5.589 make_preconditioner 11 8.9 0.000 0.000 5.567 5.589 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.643 5.360 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.146 5.300 make_m2s 4110 13.4 0.058 0.061 4.152 4.628 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.535 4.555 make_images 4110 14.4 0.398 0.445 4.041 4.514 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.107 4.107 pw_transfer 1331 11.6 0.067 0.075 3.649 3.788 density_rs2pw 110 9.6 0.004 0.005 3.290 3.712 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.542 3.682 cp_fm_diag_elpa 48 13.0 0.000 0.001 3.545 3.557 cp_fm_diag_elpa_base 48 14.0 3.498 3.519 3.542 3.554 wfi_extrapolate 11 7.9 0.001 0.001 3.530 3.530 rs_pw_transfer 902 11.9 0.012 0.014 2.957 3.425 grid_integrate_task_list 110 12.3 3.159 3.335 3.159 3.335 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.213 3.216 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.142 3.211 fft_wrap_pw1pw2_140 451 13.1 0.519 0.534 3.032 3.176 calculate_dm_sparse 110 9.5 0.001 0.001 2.953 2.985 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.892 2.917 hybrid_alltoall_any 4261 16.3 0.102 0.445 2.103 2.855 make_images_data 4110 15.4 0.046 0.050 2.353 2.839 fft3d_ps 1111 14.6 1.116 1.336 2.518 2.665 calculate_first_density_matrix 1 7.0 0.001 0.005 2.567 2.570 cp_fm_cholesky_invert 11 10.9 2.561 2.568 2.561 2.568 grid_collocate_task_list 110 9.6 2.057 2.486 2.057 2.486 potential_pw2rs 110 12.3 0.008 0.009 2.407 2.424 mp_sum_l 6514 12.8 1.720 2.404 1.720 2.404 jit_kernel_multiply 12 16.4 1.021 2.278 1.021 2.278 mp_alltoall_d11v 2046 13.8 1.793 2.051 1.793 2.051 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.943 1.962 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.875 1.876 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.812 1.822 mp_waitany 10164 13.8 1.225 1.718 1.225 1.718 multiply_cannon_metrocomm4 22605 15.4 0.076 0.079 0.768 1.681 mp_allgather_i34 2055 14.4 0.614 1.652 0.614 1.652 acc_transpose_blocks 24660 15.4 0.111 0.113 1.534 1.554 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.531 1.545 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.369 1.477 mp_irecv_dv 57340 16.2 0.642 1.464 0.642 1.464 rs_pw_transfer_RS2PW_140 121 11.5 0.205 0.215 0.967 1.434 dbcsr_complete_redistribute 325 12.2 0.242 0.303 1.140 1.407 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.018000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=557.545455, yerr=7.808662 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 658.833408E+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 9694 567. 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 0.013 0.030 63.193 63.194 qs_mol_dyn_low 1 2.0 0.003 0.003 62.926 62.936 qs_forces 11 3.9 0.003 0.003 62.856 62.858 qs_energies 11 4.9 0.001 0.001 59.643 59.648 scf_env_do_scf 11 5.9 0.000 0.001 48.485 48.485 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.022 40.023 velocity_verlet 10 3.0 0.001 0.001 32.636 32.638 dbcsr_multiply_generic 2055 12.4 0.106 0.112 31.713 32.005 multiply_cannon 2055 13.4 0.212 0.224 25.456 26.529 multiply_cannon_loop 2055 14.4 0.619 0.633 24.251 25.333 qs_scf_new_mos 99 7.5 0.001 0.001 24.891 25.011 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.890 25.010 ot_scf_mini 99 9.5 0.002 0.003 23.666 23.804 ot_mini 99 10.5 0.001 0.001 13.484 13.624 rebuild_ks_matrix 110 8.3 0.000 0.000 12.459 12.630 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.459 12.630 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.963 11.116 multiply_cannon_multrec 16440 15.4 3.793 4.674 9.573 10.469 mp_waitall_1 139946 16.5 7.172 10.399 7.172 10.399 qs_ot_get_derivative 99 11.5 0.001 0.001 8.956 9.090 init_scf_run 11 5.9 0.000 0.001 8.761 8.761 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.760 8.761 init_scf_loop 11 6.9 0.000 0.000 8.428 8.429 multiply_cannon_metrocomm3 16440 15.4 0.042 0.046 4.521 7.551 prepare_preconditioner 11 7.9 0.000 0.000 6.654 6.675 make_preconditioner 11 8.9 0.000 0.000 6.654 6.675 sum_up_and_integrate 110 10.3 0.059 0.061 6.569 6.585 integrate_v_rspace 110 11.3 0.003 0.003 6.510 6.525 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.990 6.342 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 6.036 6.037 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.997 6.008 calculate_rho_elec 110 8.6 0.059 0.059 5.996 6.007 dbcsr_mm_accdrv_process 34862 16.1 4.755 5.381 5.633 5.794 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.698 5.755 calculate_first_density_matrix 1 7.0 0.000 0.000 5.707 5.708 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.036 5.520 apply_single 110 13.6 0.000 0.000 5.036 5.520 qs_ot_get_p 110 10.4 0.001 0.001 5.349 5.506 make_m2s 4110 13.4 0.050 0.051 4.157 4.535 acc_transpose_blocks 16440 15.4 0.072 0.075 4.486 4.503 ot_diis_step 99 11.5 0.011 0.011 4.499 4.499 make_images 4110 14.4 0.390 0.507 4.041 4.419 density_rs2pw 110 9.6 0.004 0.005 3.094 4.334 multiply_cannon_sync_h2d 16440 15.4 3.656 4.207 3.656 4.207 acc_transpose_blocks_kernels 16440 16.4 0.209 0.213 4.057 4.065 rs_pw_transfer 902 11.9 0.010 0.011 2.658 3.891 jit_kernel_transpose 5 15.6 3.849 3.854 3.849 3.854 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.758 3.761 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.066 3.719 pw_transfer 1331 11.6 0.065 0.073 3.656 3.664 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.549 3.559 grid_integrate_task_list 110 12.3 3.197 3.435 3.197 3.435 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.422 3.422 fft_wrap_pw1pw2_140 451 13.1 0.635 0.642 3.067 3.080 wfi_extrapolate 11 7.9 0.001 0.001 2.970 2.970 make_images_data 4110 15.4 0.043 0.047 2.445 2.951 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.891 2.899 cp_fm_diag_elpa_base 48 14.0 2.825 2.856 2.889 2.897 hybrid_alltoall_any 4261 16.3 0.106 0.377 2.192 2.786 cp_fm_cholesky_invert 11 10.9 2.634 2.639 2.634 2.639 calculate_dm_sparse 110 9.5 0.001 0.001 2.587 2.618 multiply_cannon_metrocomm4 14385 15.4 0.046 0.049 0.910 2.531 mp_waitany 17072 13.8 1.184 2.480 1.184 2.480 grid_collocate_task_list 110 9.6 2.080 2.456 2.080 2.456 mp_irecv_dv 48980 15.7 0.839 2.407 0.839 2.407 fft3d_ps 1111 14.6 1.094 1.106 2.376 2.386 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.304 2.368 potential_pw2rs 110 12.3 0.011 0.011 2.274 2.280 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.181 0.894 2.129 mp_sum_l 6514 12.8 1.482 2.124 1.482 2.124 mp_alltoall_d11v 2046 13.8 1.730 2.058 1.730 2.058 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.953 1.954 dbcsr_complete_redistribute 325 12.2 0.346 0.367 1.406 1.862 cp_fm_upper_to_full 70 14.2 1.400 1.781 1.400 1.781 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.689 1.707 mp_allgather_i34 2055 14.4 0.530 1.661 0.530 1.661 cp_fm_cholesky_decompose 22 10.9 1.555 1.579 1.555 1.579 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.544 1.556 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.450 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.975 1.433 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.268 1.276 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.194000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.818182, yerr=8.155513 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 726.343680E+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 9694 608. 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 0.011 0.029 67.345 67.346 qs_mol_dyn_low 1 2.0 0.003 0.003 67.119 67.128 qs_forces 11 3.9 0.003 0.003 67.048 67.050 qs_energies 11 4.9 0.001 0.001 63.628 63.633 scf_env_do_scf 11 5.9 0.000 0.001 52.899 52.901 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.414 41.415 velocity_verlet 10 3.0 0.001 0.001 36.760 36.762 dbcsr_multiply_generic 2055 12.4 0.114 0.120 32.029 32.263 qs_scf_new_mos 99 7.5 0.001 0.001 26.382 26.511 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.381 26.511 multiply_cannon 2055 13.4 0.245 0.263 25.156 26.325 ot_scf_mini 99 9.5 0.002 0.003 24.756 24.871 multiply_cannon_loop 2055 14.4 0.890 0.911 23.743 24.379 ot_mini 99 10.5 0.001 0.001 14.172 14.307 multiply_cannon_multrec 24660 15.4 4.230 6.829 12.652 13.843 rebuild_ks_matrix 110 8.3 0.000 0.000 12.214 12.314 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.213 12.314 init_scf_loop 11 6.9 0.000 0.000 11.446 11.446 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.779 10.868 qs_ot_get_derivative 99 11.5 0.001 0.001 10.020 10.138 prepare_preconditioner 11 7.9 0.000 0.000 9.710 9.725 make_preconditioner 11 8.9 0.000 0.000 9.710 9.725 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.937 9.406 dbcsr_mm_accdrv_process 52304 16.0 6.999 8.669 8.276 9.205 init_scf_run 11 5.9 0.000 0.001 8.010 8.010 scf_env_initial_rho_setup 11 6.9 0.000 0.001 8.010 8.010 sum_up_and_integrate 110 10.3 0.067 0.070 6.523 6.536 integrate_v_rspace 110 11.3 0.003 0.003 6.456 6.468 mp_waitall_1 121746 16.5 4.375 6.445 4.375 6.445 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.079 6.091 calculate_rho_elec 110 8.6 0.078 0.081 6.079 6.090 qs_ot_get_p 110 10.4 0.001 0.001 5.604 5.751 make_m2s 4110 13.4 0.060 0.063 5.309 5.661 make_images 4110 14.4 0.575 0.698 5.166 5.515 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 5.360 5.362 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.021 5.054 calculate_first_density_matrix 1 7.0 0.000 0.000 4.966 4.969 cp_fm_upper_to_full 70 14.2 3.364 4.879 3.364 4.879 acc_transpose_blocks 24660 15.4 0.107 0.110 4.125 4.152 ot_diis_step 99 11.5 0.011 0.011 4.115 4.115 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.032 4.112 apply_single 110 13.6 0.000 0.000 4.032 4.112 qs_ot_p2m_diag 48 11.0 0.055 0.063 3.843 3.857 pw_transfer 1331 11.6 0.066 0.076 3.755 3.785 dbcsr_complete_redistribute 325 12.2 0.414 0.468 2.608 3.753 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.649 3.683 density_rs2pw 110 9.6 0.004 0.004 3.015 3.531 acc_transpose_blocks_kernels 24660 16.4 0.305 0.309 3.470 3.485 grid_integrate_task_list 110 12.3 3.262 3.430 3.262 3.430 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.355 3.355 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.259 3.305 multiply_cannon_sync_h2d 24660 15.4 3.172 3.284 3.172 3.284 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.106 3.237 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.569 3.236 jit_kernel_transpose 5 15.6 3.165 3.176 3.165 3.176 fft_wrap_pw1pw2_140 451 13.1 0.665 0.685 3.126 3.164 hybrid_alltoall_any 4261 16.3 0.120 0.456 2.295 3.067 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.982 3.051 make_images_data 4110 15.4 0.046 0.050 2.688 3.032 wfi_extrapolate 11 7.9 0.001 0.001 2.953 2.953 calculate_dm_sparse 110 9.5 0.001 0.001 2.908 2.938 rs_pw_transfer 902 11.9 0.010 0.010 2.366 2.891 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.764 2.881 mp_alltoall_i22 605 13.7 1.651 2.845 1.651 2.845 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.821 2.830 cp_fm_diag_elpa_base 48 14.0 2.673 2.733 2.819 2.829 cp_fm_cholesky_invert 11 10.9 2.606 2.615 2.606 2.615 grid_collocate_task_list 110 9.6 2.181 2.444 2.181 2.444 fft3d_ps 1111 14.6 1.085 1.118 2.415 2.428 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.235 2.238 potential_pw2rs 110 12.3 0.012 0.013 2.177 2.185 mp_alltoall_d11v 2046 13.8 1.721 2.070 1.721 2.070 jit_kernel_multiply 10 15.9 0.947 1.820 0.947 1.820 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.695 1.729 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.598 1.701 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.674 1.685 cp_fm_cholesky_decompose 22 10.9 1.587 1.624 1.587 1.624 mp_allgather_i34 2055 14.4 0.515 1.595 0.515 1.595 mp_waitany 13376 13.8 1.049 1.572 1.049 1.572 mp_sum_l 6514 12.8 0.876 1.559 0.876 1.559 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.522 1.541 multiply_cannon_metrocomm4 20550 15.4 0.062 0.065 0.848 1.478 mp_irecv_dv 62702 16.1 0.746 1.400 0.746 1.400 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.346000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=687.636364, yerr=6.636986 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 830.382080E+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 9694 649. 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 0.028 55.461 55.463 qs_mol_dyn_low 1 2.0 0.003 0.003 55.187 55.196 qs_forces 11 3.9 0.020 0.025 55.115 55.117 qs_energies 11 4.9 0.001 0.001 51.391 51.402 scf_env_do_scf 11 5.9 0.000 0.001 43.162 43.162 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.367 35.367 velocity_verlet 10 3.0 0.001 0.001 31.155 31.158 dbcsr_multiply_generic 2055 12.4 0.104 0.109 22.928 23.037 qs_scf_new_mos 99 7.5 0.001 0.001 20.297 20.357 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.296 20.356 ot_scf_mini 99 9.5 0.002 0.002 19.061 19.119 multiply_cannon 2055 13.4 0.243 0.252 17.451 18.764 multiply_cannon_loop 2055 14.4 0.322 0.337 16.120 16.349 rebuild_ks_matrix 110 8.3 0.000 0.000 11.967 11.999 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.966 11.999 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.611 10.639 ot_mini 99 10.5 0.001 0.001 10.354 10.404 multiply_cannon_multrec 8220 15.4 3.257 4.694 7.597 8.866 init_scf_loop 11 6.9 0.000 0.000 7.749 7.750 mp_waitall_1 103326 16.6 6.080 7.733 6.080 7.733 qs_ot_get_derivative 99 11.5 0.001 0.001 6.524 6.580 sum_up_and_integrate 110 10.3 0.079 0.081 6.551 6.560 integrate_v_rspace 110 11.3 0.004 0.012 6.472 6.480 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.278 6.292 calculate_rho_elec 110 8.6 0.114 0.115 6.277 6.292 prepare_preconditioner 11 7.9 0.000 0.000 6.040 6.048 make_preconditioner 11 8.9 0.000 0.000 6.040 6.048 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.618 5.689 dbcsr_mm_accdrv_process 17442 15.9 2.839 3.849 4.209 5.189 init_scf_run 11 5.9 0.000 0.001 5.106 5.106 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.106 5.106 qs_ot_get_p 110 10.4 0.001 0.001 4.810 4.873 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.146 4.567 make_m2s 4110 13.4 0.039 0.040 4.220 4.470 make_images 4110 14.4 0.635 0.687 4.089 4.338 pw_transfer 1331 11.6 0.066 0.072 3.939 3.955 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.831 3.852 ot_diis_step 99 11.5 0.012 0.012 3.810 3.810 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.727 3.769 apply_single 110 13.6 0.000 0.000 3.727 3.768 grid_integrate_task_list 110 12.3 3.365 3.540 3.365 3.540 density_rs2pw 110 9.6 0.004 0.004 3.027 3.399 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.368 3.372 fft_wrap_pw1pw2_140 451 13.1 0.831 0.843 3.300 3.323 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.061 3.061 multiply_cannon_sync_h2d 8220 15.4 2.900 3.030 2.900 3.030 cp_fm_cholesky_invert 11 10.9 2.906 2.910 2.906 2.910 make_images_data 4110 15.4 0.038 0.044 2.377 2.791 wfi_extrapolate 11 7.9 0.001 0.001 2.725 2.725 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.669 2.677 hybrid_alltoall_any 4261 16.3 0.201 0.864 2.304 2.673 rs_pw_transfer 902 11.9 0.010 0.010 2.196 2.601 cp_fm_diag_elpa 48 13.0 0.001 0.001 2.544 2.552 cp_fm_diag_elpa_base 48 14.0 2.488 2.512 2.542 2.549 calculate_dm_sparse 110 9.5 0.001 0.001 2.506 2.548 grid_collocate_task_list 110 9.6 2.277 2.537 2.277 2.537 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.526 2.527 fft3d_ps 1111 14.6 1.134 1.175 2.381 2.397 calculate_first_density_matrix 1 7.0 0.000 0.000 2.286 2.287 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.184 2.196 potential_pw2rs 110 12.3 0.015 0.015 2.091 2.100 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.775 1.993 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.863 1.890 mp_alltoall_d11v 2046 13.8 1.616 1.869 1.616 1.869 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.762 1.790 cp_fm_cholesky_decompose 22 10.9 1.697 1.717 1.697 1.717 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.684 1.688 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.502 1.624 mp_allgather_i34 2055 14.4 0.519 1.610 0.519 1.610 dbcsr_complete_redistribute 325 12.2 0.559 0.579 1.438 1.532 mp_waitany 9240 13.8 1.067 1.502 1.067 1.502 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.477 1.488 jit_kernel_multiply 8 15.8 1.061 1.477 1.061 1.477 qs_create_task_list 11 7.9 0.000 0.000 1.217 1.324 generate_qs_task_list 11 8.9 0.374 0.440 1.217 1.324 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.475 1.243 rs_pw_transfer_RS2PW_140 121 11.5 0.160 0.164 0.794 1.204 mp_irecv_dv 24056 15.7 0.451 1.201 0.451 1.201 rs_gather_matrices 110 12.3 0.327 0.369 0.924 1.164 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.115 1.133 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.463000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=779.818182, yerr=11.983460 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 1.336275E+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 9672 819. 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 0.017 0.037 88.531 88.532 qs_mol_dyn_low 1 2.0 0.003 0.003 88.231 88.241 qs_forces 11 3.9 0.003 0.003 88.159 88.160 qs_energies 11 4.9 0.001 0.001 84.009 84.010 scf_env_do_scf 11 5.9 0.001 0.001 72.928 72.928 velocity_verlet 10 3.0 0.001 0.001 56.076 56.082 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 44.294 44.296 dbcsr_multiply_generic 2055 12.4 0.119 0.124 29.677 29.800 init_scf_loop 11 6.9 0.000 0.000 28.563 28.565 prepare_preconditioner 11 7.9 0.000 0.000 26.527 26.542 make_preconditioner 11 8.9 0.000 0.000 26.527 26.542 qs_scf_new_mos 99 7.5 0.001 0.001 26.070 26.164 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.069 26.164 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.708 25.998 ot_scf_mini 99 9.5 0.002 0.002 24.320 24.406 multiply_cannon 2055 13.4 0.346 0.372 22.662 23.466 multiply_cannon_loop 2055 14.4 0.341 0.344 20.859 21.167 cp_fm_upper_to_full 70 14.2 12.756 18.290 12.756 18.290 rebuild_ks_matrix 110 8.3 0.001 0.001 14.137 14.217 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.136 14.217 ot_mini 99 10.5 0.001 0.001 13.566 13.647 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.794 12.866 dbcsr_complete_redistribute 325 12.2 1.027 1.066 7.366 10.601 multiply_cannon_multrec 8220 15.4 4.364 4.567 9.565 9.670 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.356 9.594 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.804 9.021 qs_ot_get_derivative 99 11.5 0.001 0.001 8.894 8.979 mp_alltoall_i22 605 13.7 5.435 8.649 5.435 8.649 mp_waitall_1 84994 16.7 7.557 8.527 7.557 8.527 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.673 7.712 calculate_rho_elec 110 8.6 0.224 0.225 7.672 7.711 sum_up_and_integrate 110 10.3 0.151 0.152 7.347 7.365 integrate_v_rspace 110 11.3 0.004 0.004 7.196 7.214 init_scf_run 11 5.9 0.000 0.001 6.828 6.828 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.828 6.828 make_m2s 4110 13.4 0.044 0.044 5.221 5.760 qs_ot_get_p 110 10.4 0.001 0.001 5.572 5.663 make_images 4110 14.4 0.873 0.923 5.032 5.569 cp_fm_cholesky_invert 11 10.9 5.407 5.412 5.407 5.412 dbcsr_mm_accdrv_process 11614 15.7 3.317 3.664 5.058 5.301 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.765 5.289 apply_single 110 13.6 0.000 0.000 4.765 5.288 pw_transfer 1331 11.6 0.075 0.076 5.043 5.049 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.791 5.032 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.926 4.933 ot_diis_step 99 11.5 0.015 0.016 4.653 4.653 fft_wrap_pw1pw2_140 451 13.1 1.334 1.338 4.291 4.299 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.003 4.008 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.000 4.007 multiply_cannon_sync_h2d 8220 15.4 3.943 3.951 3.943 3.951 density_rs2pw 110 9.6 0.004 0.004 3.707 3.733 grid_integrate_task_list 110 12.3 3.667 3.731 3.667 3.731 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.664 3.665 hybrid_alltoall_any 4261 16.3 0.257 0.553 2.863 3.614 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.561 3.562 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.507 3.534 make_images_data 4110 15.4 0.042 0.045 2.866 3.509 calculate_first_density_matrix 1 7.0 0.000 0.000 3.389 3.390 wfi_extrapolate 11 7.9 0.001 0.001 3.317 3.317 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.821 3.290 calculate_dm_sparse 110 9.5 0.001 0.001 3.133 3.152 cp_fm_diag_elpa 48 13.0 0.001 0.001 2.963 2.964 cp_fm_diag_elpa_base 48 14.0 2.425 2.611 2.961 2.961 fft3d_ps 1111 14.6 1.303 1.308 2.861 2.869 grid_collocate_task_list 110 9.6 2.636 2.664 2.636 2.664 potential_pw2rs 110 12.3 0.021 0.021 2.439 2.444 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.373 2.417 rs_pw_transfer 902 11.9 0.010 0.011 2.284 2.301 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.190 2.240 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.078 2.147 cp_fm_cholesky_decompose 22 10.9 2.028 2.051 2.028 2.051 acc_transpose_blocks 8220 15.4 0.035 0.035 2.040 2.044 mp_alltoall_d11v 2046 13.8 1.946 2.031 1.946 2.031 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.942 1.951 qs_create_task_list 11 7.9 0.000 0.000 1.891 1.939 generate_qs_task_list 11 8.9 0.730 0.785 1.891 1.938 acc_transpose_blocks_kernels 8220 16.4 0.110 0.111 1.800 1.806 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.752 1.795 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.532000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1198.272727, yerr=58.939246 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 629.608448E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57767. MP_Allreduce 11004 802. MP_Sync 87 MP_Alltoall 2226 3242816. 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 0.019 0.039 208.133 208.134 qs_mol_dyn_low 1 2.0 0.003 0.003 207.758 207.780 qs_forces 11 3.9 0.004 0.005 207.667 207.669 qs_energies 11 4.9 0.001 0.001 202.024 202.039 scf_env_do_scf 11 5.9 0.001 0.001 185.439 185.443 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 164.764 164.765 dbcsr_multiply_generic 2507 12.6 0.182 0.186 125.251 126.143 qs_scf_new_mos 117 7.6 0.001 0.001 124.649 124.963 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.649 124.963 velocity_verlet 10 3.0 0.001 0.001 124.293 124.295 ot_scf_mini 117 9.6 0.003 0.003 118.036 118.354 multiply_cannon 2507 13.6 0.239 0.249 101.464 103.206 multiply_cannon_loop 2507 14.6 2.084 2.132 99.298 101.154 ot_mini 117 10.6 0.001 0.001 66.378 66.666 multiply_cannon_multrec 60168 15.6 33.359 34.923 41.578 43.387 qs_ot_get_derivative 117 11.6 0.001 0.001 41.458 41.769 rebuild_ks_matrix 128 8.3 0.001 0.001 33.765 34.168 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.765 34.167 mp_waitall_1 267128 16.5 28.747 32.433 28.747 32.433 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.329 30.713 qs_ot_get_p 128 10.4 0.001 0.001 29.600 29.910 multiply_cannon_sync_h2d 60168 15.6 27.625 29.653 27.625 29.653 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.502 25.527 apply_single 128 13.6 0.001 0.001 24.502 25.527 ot_diis_step 117 11.6 0.008 0.008 24.695 24.696 qs_ot_p2m_diag 83 11.4 0.079 0.091 23.041 23.126 init_scf_loop 11 6.9 0.000 0.001 20.599 20.600 cp_dbcsr_syevd 83 12.4 0.004 0.005 20.386 20.387 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.908 20.113 multiply_cannon_metrocomm3 60168 15.6 0.113 0.118 16.148 18.638 cp_fm_diag_elpa 83 13.4 0.001 0.001 17.411 17.445 cp_fm_diag_elpa_base 83 14.4 17.344 17.376 17.407 17.439 prepare_preconditioner 11 7.9 0.000 0.000 15.996 16.031 make_preconditioner 11 8.9 0.000 0.000 15.996 16.031 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.214 15.359 sum_up_and_integrate 128 10.3 0.088 0.107 14.254 14.268 make_m2s 5014 13.6 0.107 0.119 13.815 14.192 integrate_v_rspace 128 11.3 0.004 0.005 14.165 14.184 make_images 5014 14.6 0.397 0.416 13.629 14.019 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.763 13.896 calculate_rho_elec 128 8.7 0.046 0.065 13.762 13.896 init_scf_run 11 5.9 0.000 0.001 12.374 12.374 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.374 12.374 density_rs2pw 128 9.7 0.006 0.007 7.097 10.376 mp_sum_l 7870 13.0 8.690 9.811 8.690 9.811 wfi_extrapolate 11 7.9 0.001 0.001 9.192 9.192 cp_fm_cholesky_invert 11 10.9 9.158 9.166 9.158 9.166 rs_pw_transfer 1046 11.9 0.016 0.018 5.837 9.087 calculate_dm_sparse 128 9.5 0.001 0.001 8.417 8.563 dbcsr_mm_accdrv_process 124484 16.2 3.175 3.282 7.779 8.269 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.065 8.220 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.001 8.118 multiply_cannon_metrocomm1 60168 15.6 0.088 0.092 6.016 7.962 pw_transfer 1547 11.6 0.074 0.091 7.774 7.957 make_images_data 5014 15.6 0.063 0.069 6.805 7.766 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.571 7.756 grid_integrate_task_list 128 12.3 7.056 7.621 7.056 7.621 hybrid_alltoall_any 5200 16.5 0.294 2.260 5.947 7.256 fft_wrap_pw1pw2_140 523 13.2 1.272 1.321 6.648 6.828 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.650 6.659 mp_waitany 16020 13.9 2.735 5.968 2.735 5.968 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.813 5.931 grid_collocate_task_list 128 9.7 4.556 5.785 4.556 5.785 fft3d_ps 1291 14.7 2.146 2.780 5.342 5.667 mp_alltoall_d11v 2415 14.1 4.298 5.600 4.298 5.600 rs_pw_transfer_RS2PW_140 139 11.5 0.277 0.296 2.147 5.421 cp_fm_cholesky_decompose 22 10.9 4.655 4.669 4.655 4.669 potential_pw2rs 128 12.3 0.009 0.011 4.631 4.660 mp_sum_d 4470 12.1 3.446 4.283 3.446 4.283 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.134000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.181818, yerr=6.685373 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 829.124608E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58363. MP_Allreduce 10978 967. MP_Sync 87 MP_Alltoall 1969 5007368. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.050 193.961 193.963 qs_mol_dyn_low 1 2.0 0.003 0.003 193.566 193.581 qs_forces 11 3.9 0.010 0.019 193.479 193.480 qs_energies 11 4.9 0.001 0.002 186.676 186.689 scf_env_do_scf 11 5.9 0.001 0.001 167.070 167.080 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 134.043 134.045 velocity_verlet 10 3.0 0.001 0.001 120.089 120.090 dbcsr_multiply_generic 2507 12.6 0.188 0.193 100.646 102.000 qs_scf_new_mos 117 7.6 0.001 0.001 93.991 94.453 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.990 94.452 ot_scf_mini 117 9.6 0.004 0.004 89.181 89.720 multiply_cannon 2507 13.6 0.477 0.529 80.575 84.845 multiply_cannon_loop 2507 14.6 1.256 1.299 77.273 80.128 ot_mini 117 10.6 0.001 0.001 49.825 50.294 mp_waitall_1 214728 16.6 24.756 38.924 24.756 38.924 multiply_cannon_multrec 30084 15.6 21.861 26.088 31.449 36.158 rebuild_ks_matrix 128 8.3 0.001 0.001 33.297 33.912 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 33.296 33.912 init_scf_loop 11 6.9 0.000 0.000 32.939 32.939 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.953 30.501 multiply_cannon_metrocomm3 30084 15.6 0.090 0.095 15.879 28.891 prepare_preconditioner 11 7.9 0.000 0.000 28.552 28.630 make_preconditioner 11 8.9 0.000 0.000 28.552 28.630 qs_ot_get_derivative 117 11.6 0.001 0.002 27.734 28.262 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.212 27.739 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.200 23.323 apply_single 128 13.6 0.001 0.001 22.200 23.323 ot_diis_step 117 11.6 0.014 0.015 21.923 21.925 qs_ot_get_p 128 10.4 0.001 0.001 21.141 21.729 multiply_cannon_sync_h2d 30084 15.6 19.127 21.534 19.127 21.534 cp_fm_cholesky_invert 11 10.9 16.655 16.668 16.655 16.668 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.370 16.405 make_m2s 5014 13.6 0.091 0.097 14.339 15.873 make_images 5014 14.6 1.149 1.331 14.125 15.655 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.231 15.232 init_scf_run 11 5.9 0.000 0.001 14.818 14.820 scf_env_initial_rho_setup 11 6.9 0.001 0.001 14.818 14.819 sum_up_and_integrate 128 10.3 0.115 0.130 14.709 14.741 integrate_v_rspace 128 11.3 0.004 0.005 14.594 14.628 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.069 14.106 calculate_rho_elec 128 8.7 0.088 0.104 14.068 14.105 cp_fm_diag_elpa 83 13.4 0.001 0.001 12.054 12.081 cp_fm_diag_elpa_base 83 14.4 11.799 11.917 12.048 12.077 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.916 11.275 multiply_cannon_metrocomm4 27577 15.6 0.100 0.113 3.790 10.781 make_images_data 5014 15.6 0.063 0.071 8.780 10.634 mp_irecv_dv 69486 16.3 3.592 10.395 3.592 10.395 density_rs2pw 128 9.7 0.006 0.007 7.426 10.047 hybrid_alltoall_any 5200 16.5 0.343 1.510 7.566 9.860 dbcsr_mm_accdrv_process 62242 16.2 4.471 5.155 9.052 9.640 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 9.443 9.452 pw_transfer 1547 11.6 0.086 0.096 8.811 8.869 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 8.570 8.717 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 8.586 8.648 wfi_extrapolate 11 7.9 0.001 0.001 8.399 8.399 rs_pw_transfer 1046 11.9 0.014 0.017 5.728 8.323 grid_integrate_task_list 128 12.3 7.184 7.689 7.184 7.689 fft_wrap_pw1pw2_140 523 13.2 1.333 1.353 7.569 7.650 cp_fm_cholesky_decompose 22 10.9 6.907 6.998 6.907 6.998 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.175 6.939 calculate_dm_sparse 128 9.5 0.001 0.001 6.507 6.636 calculate_first_density_matrix 1 7.0 0.000 0.001 6.238 6.245 fft3d_ps 1291 14.7 2.815 2.990 5.901 5.944 mp_sum_l 7870 13.0 3.950 5.901 3.950 5.901 grid_collocate_task_list 128 9.7 4.706 5.885 4.706 5.885 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.350 5.414 acc_transpose_blocks 30084 15.6 0.140 0.145 5.120 5.145 mp_waitany 11748 13.9 2.496 5.107 2.496 5.107 mp_allgather_i34 2507 14.6 1.877 5.107 1.877 5.107 potential_pw2rs 128 12.3 0.016 0.019 4.979 5.001 mp_alltoall_d11v 2415 14.1 4.194 4.840 4.194 4.840 rs_pw_transfer_RS2PW_140 139 11.5 0.349 0.375 2.093 4.677 acc_transpose_blocks_kernels 30084 16.6 0.396 0.404 4.289 4.301 dbcsr_complete_redistribute 395 12.7 0.769 0.852 3.203 4.047 mp_sum_d 4465 12.1 2.606 3.937 2.606 3.937 jit_kernel_transpose 5 15.6 3.893 3.902 3.893 3.902 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=193.963000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.818182, yerr=2.587318 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.950886E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 944.349184E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889764224 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57903. MP_Allreduce 11057 1004. MP_Sync 87 MP_Alltoall 1724 9394172. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.032 176.442 176.443 qs_mol_dyn_low 1 2.0 0.003 0.003 176.035 176.047 qs_forces 11 3.9 0.004 0.005 175.905 175.910 qs_energies 11 4.9 0.001 0.001 169.239 169.250 scf_env_do_scf 11 5.9 0.001 0.001 153.263 153.264 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 118.413 118.414 velocity_verlet 10 3.0 0.001 0.001 113.538 113.541 dbcsr_multiply_generic 2527 12.6 0.184 0.191 82.715 83.756 qs_scf_new_mos 118 7.6 0.001 0.001 79.835 80.177 qs_scf_loop_do_ot 118 8.6 0.001 0.001 79.834 80.176 ot_scf_mini 118 9.6 0.003 0.004 75.652 76.031 multiply_cannon 2527 13.6 0.505 0.523 62.990 67.422 multiply_cannon_loop 2527 14.6 0.865 0.896 59.774 62.001 ot_mini 118 10.6 0.001 0.001 42.809 43.188 init_scf_loop 11 6.9 0.000 0.000 34.753 34.754 mp_waitall_1 171878 16.6 25.424 34.674 25.424 34.674 rebuild_ks_matrix 129 8.3 0.001 0.001 31.094 31.588 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 31.093 31.587 prepare_preconditioner 11 7.9 0.000 0.000 30.709 30.756 make_preconditioner 11 8.9 0.000 0.000 30.709 30.756 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.372 29.772 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.988 28.439 multiply_cannon_multrec 20216 15.6 13.482 16.866 22.154 25.422 multiply_cannon_metrocomm3 20216 15.6 0.058 0.063 15.733 24.751 qs_ot_get_derivative 118 11.6 0.002 0.002 22.647 23.030 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.288 21.392 apply_single 129 13.6 0.001 0.001 20.288 21.392 ot_diis_step 118 11.6 0.018 0.025 20.052 20.052 qs_ot_get_p 129 10.4 0.001 0.001 18.048 18.503 multiply_cannon_sync_h2d 20216 15.6 14.296 16.108 14.296 16.108 make_m2s 5054 13.6 0.083 0.091 14.736 15.827 make_images 5054 14.6 1.186 1.283 14.497 15.584 sum_up_and_integrate 129 10.3 0.134 0.145 14.770 14.797 integrate_v_rspace 129 11.3 0.004 0.004 14.635 14.664 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.388 14.424 calculate_rho_elec 129 8.7 0.131 0.146 14.387 14.423 cp_fm_cholesky_invert 11 10.9 14.334 14.343 14.334 14.343 qs_ot_p2m_diag 83 11.4 0.265 0.272 13.846 13.854 cp_dbcsr_syevd 83 12.4 0.005 0.005 12.846 12.847 init_scf_run 11 5.9 0.000 0.001 10.984 10.984 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.984 10.984 make_images_data 5054 15.6 0.060 0.070 9.142 10.604 cp_fm_diag_elpa 83 13.4 0.001 0.001 9.778 9.799 cp_fm_diag_elpa_base 83 14.4 9.374 9.539 9.774 9.796 hybrid_alltoall_any 5240 16.5 0.438 2.002 7.958 9.742 multiply_cannon_metrocomm4 17689 15.6 0.064 0.074 3.535 9.584 density_rs2pw 129 9.7 0.006 0.006 7.198 9.440 mp_irecv_dv 50625 16.2 3.410 9.333 3.410 9.333 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.824 9.089 pw_transfer 1559 11.6 0.086 0.102 8.970 9.073 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 8.744 8.851 dbcsr_mm_accdrv_process 41830 16.2 4.481 5.257 8.122 8.256 grid_integrate_task_list 129 12.3 7.366 7.905 7.366 7.905 fft_wrap_pw1pw2_140 527 13.2 1.423 1.451 7.755 7.877 wfi_extrapolate 11 7.9 0.001 0.001 7.426 7.426 rs_pw_transfer 1054 12.0 0.014 0.015 5.166 7.425 cp_fm_cholesky_decompose 22 10.9 7.259 7.288 7.259 7.288 cp_fm_upper_to_full 105 14.8 5.727 7.218 5.727 7.218 dbcsr_complete_redistribute 395 12.7 1.168 1.199 4.430 6.213 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.005 6.013 fft3d_ps 1301 14.7 2.758 2.995 5.850 5.928 calculate_dm_sparse 129 9.5 0.001 0.001 5.776 5.877 grid_collocate_task_list 129 9.7 4.930 5.745 4.930 5.745 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.594 5.322 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.133 5.268 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.250 5.031 mp_alltoall_d11v 2423 14.1 4.272 4.846 4.272 4.846 mp_allgather_i34 2527 14.6 1.728 4.838 1.728 4.838 potential_pw2rs 129 12.3 0.020 0.023 4.754 4.769 mp_waitany 11836 13.9 2.388 4.698 2.388 4.698 mp_sum_l 7930 13.1 3.147 4.507 3.147 4.507 rs_pw_transfer_RS2PW_140 140 11.5 0.334 0.356 1.940 4.183 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.059 4.087 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.318 4.053 mp_alltoall_i22 718 14.1 1.902 3.786 1.902 3.786 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.745 3.746 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.607 3.653 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.443000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=888.818182, yerr=12.611814 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.153642E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58195. MP_Allreduce 11005 1091. MP_Sync 86 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.033 187.170 187.171 qs_mol_dyn_low 1 2.0 0.003 0.003 186.772 186.787 qs_forces 11 3.9 0.004 0.004 186.650 186.655 qs_energies 11 4.9 0.001 0.001 179.501 179.512 scf_env_do_scf 11 5.9 0.001 0.001 162.843 162.850 velocity_verlet 10 3.0 0.001 0.001 123.640 123.643 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.365 116.366 qs_scf_new_mos 116 7.6 0.001 0.001 79.122 79.360 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.121 79.359 dbcsr_multiply_generic 2485 12.5 0.186 0.194 78.542 79.358 ot_scf_mini 116 9.6 0.003 0.003 74.665 74.958 multiply_cannon 2485 13.5 0.550 0.582 54.450 56.632 multiply_cannon_loop 2485 14.5 1.176 1.201 50.663 52.570 init_scf_loop 11 6.9 0.000 0.000 46.358 46.359 prepare_preconditioner 11 7.9 0.000 0.000 42.262 42.291 make_preconditioner 11 8.9 0.000 0.000 42.262 42.291 ot_mini 116 10.6 0.001 0.001 41.718 41.996 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.838 40.875 multiply_cannon_multrec 29820 15.5 13.984 19.781 25.902 31.250 rebuild_ks_matrix 127 8.3 0.001 0.001 29.688 29.971 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 29.687 29.971 mp_waitall_1 146592 16.7 17.079 27.417 17.079 27.417 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.749 27.002 qs_ot_get_derivative 116 11.6 0.001 0.002 22.338 22.635 make_m2s 4970 13.5 0.097 0.101 19.815 20.813 make_images 4970 14.5 1.921 2.234 19.508 20.510 qs_ot_get_p 127 10.4 0.001 0.001 19.368 19.675 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.787 19.385 apply_single 127 13.6 0.001 0.001 18.787 19.385 ot_diis_step 116 11.6 0.017 0.018 19.263 19.264 cp_fm_upper_to_full 104 14.8 11.446 16.870 11.446 16.870 cp_fm_cholesky_invert 11 10.9 16.107 16.116 16.107 16.116 multiply_cannon_metrocomm3 29820 15.5 0.047 0.050 6.352 15.331 qs_ot_p2m_diag 82 11.4 0.338 0.385 15.276 15.330 sum_up_and_integrate 127 10.3 0.138 0.149 14.598 14.617 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.496 14.534 calculate_rho_elec 127 8.7 0.172 0.188 14.495 14.533 integrate_v_rspace 127 11.3 0.004 0.004 14.459 14.483 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.013 14.015 dbcsr_complete_redistribute 393 12.7 1.506 1.629 8.976 12.745 multiply_cannon_sync_h2d 29820 15.5 11.653 12.527 11.653 12.527 make_images_data 4970 15.5 0.062 0.068 10.706 12.286 dbcsr_mm_accdrv_process 61748 16.2 7.363 8.439 11.496 11.947 hybrid_alltoall_any 5155 16.4 0.519 2.190 9.542 11.404 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.617 11.353 cp_fm_diag_elpa 82 13.4 0.001 0.001 10.934 10.946 cp_fm_diag_elpa_base 82 14.4 9.982 10.294 10.927 10.938 init_scf_run 11 5.9 0.000 0.004 10.790 10.791 scf_env_initial_rho_setup 11 6.9 0.002 0.004 10.789 10.791 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.401 10.070 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.217 9.434 mp_alltoall_i22 712 14.1 5.617 9.396 5.617 9.396 pw_transfer 1535 11.6 0.085 0.101 9.093 9.167 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.868 8.946 density_rs2pw 127 9.7 0.006 0.006 7.003 8.529 fft_wrap_pw1pw2_140 519 13.2 1.541 1.571 7.889 7.981 grid_integrate_task_list 127 12.3 7.462 7.952 7.462 7.952 wfi_extrapolate 11 7.9 0.001 0.001 7.654 7.654 cp_fm_cholesky_decompose 22 10.9 7.538 7.617 7.538 7.617 multiply_cannon_metrocomm4 24850 15.5 0.083 0.096 2.766 6.986 mp_irecv_dv 75445 16.2 2.616 6.720 2.616 6.720 calculate_dm_sparse 127 9.5 0.001 0.001 6.225 6.323 rs_pw_transfer 1038 11.9 0.013 0.014 4.652 6.248 fft3d_ps 1281 14.7 2.833 2.903 5.831 5.898 grid_collocate_task_list 127 9.7 5.030 5.715 5.030 5.715 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.398 5.448 mp_alltoall_d11v 2401 14.1 4.761 5.259 4.761 5.259 potential_pw2rs 127 12.3 0.022 0.023 4.476 4.489 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.355 4.454 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.318 4.410 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.400 4.402 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.104 4.177 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.171000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1080.000000, yerr=18.040359 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.484722E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3991 58365. MP_Allreduce 10975 1176. MP_Sync 86 MP_Alltoall 1700 18828174. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.033 173.111 173.112 qs_mol_dyn_low 1 2.0 0.003 0.003 172.677 172.684 qs_forces 11 3.9 0.004 0.004 172.548 172.558 qs_energies 11 4.9 0.001 0.001 165.003 165.011 scf_env_do_scf 11 5.9 0.001 0.001 147.797 147.801 velocity_verlet 10 3.0 0.001 0.001 112.893 112.897 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 111.973 111.974 qs_scf_new_mos 116 7.6 0.001 0.001 73.771 73.857 qs_scf_loop_do_ot 116 8.6 0.001 0.001 73.770 73.856 dbcsr_multiply_generic 2485 12.5 0.179 0.190 73.370 73.854 ot_scf_mini 116 9.6 0.003 0.004 69.372 69.426 multiply_cannon 2485 13.5 0.579 0.616 53.875 57.980 multiply_cannon_loop 2485 14.5 0.441 0.451 49.297 50.137 ot_mini 116 10.6 0.001 0.001 39.042 39.094 init_scf_loop 11 6.9 0.000 0.000 35.678 35.679 mp_waitall_1 124680 16.7 25.575 32.893 25.575 32.893 prepare_preconditioner 11 7.9 0.000 0.000 31.692 31.730 make_preconditioner 11 8.9 0.000 0.000 31.692 31.730 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.627 29.873 rebuild_ks_matrix 127 8.3 0.001 0.001 29.689 29.769 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 29.688 29.768 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.945 27.021 multiply_cannon_multrec 9940 15.5 10.294 14.513 17.626 20.672 ot_diis_step 116 11.6 0.020 0.021 19.665 19.665 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.176 19.559 apply_single 127 13.6 0.001 0.001 19.176 19.558 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 12.141 19.371 qs_ot_get_derivative 116 11.6 0.001 0.002 19.309 19.362 cp_fm_cholesky_invert 11 10.9 18.157 18.163 18.157 18.163 make_m2s 4970 13.5 0.067 0.071 15.974 18.011 make_images 4970 14.5 2.253 2.766 15.667 17.704 qs_ot_get_p 127 10.4 0.001 0.001 17.539 17.616 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.191 15.219 calculate_rho_elec 127 8.7 0.254 0.265 15.190 15.219 sum_up_and_integrate 127 10.3 0.178 0.188 14.910 14.966 integrate_v_rspace 127 11.3 0.004 0.004 14.732 14.796 qs_ot_p2m_diag 82 11.4 0.489 0.495 13.869 13.886 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.750 12.752 make_images_data 4970 15.5 0.051 0.061 9.777 12.108 hybrid_alltoall_any 5155 16.4 0.811 3.598 9.574 12.074 multiply_cannon_sync_h2d 9940 15.5 11.559 11.980 11.559 11.980 init_scf_run 11 5.9 0.000 0.003 10.440 10.440 scf_env_initial_rho_setup 11 6.9 0.005 0.011 10.439 10.440 cp_fm_diag_elpa 82 13.4 0.001 0.001 9.724 9.736 cp_fm_diag_elpa_base 82 14.4 9.479 9.562 9.716 9.728 pw_transfer 1535 11.6 0.085 0.095 9.445 9.472 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.223 9.256 fft_wrap_pw1pw2_140 519 13.2 1.890 1.921 8.168 8.201 grid_integrate_task_list 127 12.3 7.753 8.103 7.753 8.103 cp_fm_cholesky_decompose 22 10.9 7.941 8.064 7.941 8.064 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.721 7.764 density_rs2pw 127 9.7 0.006 0.006 6.825 7.700 dbcsr_mm_accdrv_process 20590 16.1 2.656 3.401 6.963 7.588 wfi_extrapolate 11 7.9 0.001 0.001 7.412 7.412 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.381 7.181 mp_allgather_i34 2485 14.5 2.801 7.087 2.801 7.087 calculate_dm_sparse 127 9.5 0.001 0.001 6.163 6.233 grid_collocate_task_list 127 9.7 5.350 6.015 5.350 6.015 fft3d_ps 1281 14.7 2.745 2.821 5.743 5.767 mp_alltoall_d11v 2401 14.1 4.902 5.462 4.902 5.462 dbcsr_complete_redistribute 393 12.7 2.118 2.161 5.089 5.439 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.278 5.278 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.132 5.138 rs_pw_transfer 1038 11.9 0.012 0.013 4.077 4.973 potential_pw2rs 127 12.3 0.026 0.026 4.399 4.416 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.063 4.094 multiply_cannon_metrocomm4 7455 15.5 0.024 0.026 1.829 3.969 mp_irecv_dv 28618 15.9 1.793 3.897 1.793 3.897 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.571 3.846 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.654 3.680 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.640 3.648 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.312 3.607 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.521 3.590 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.112000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1392.454545, yerr=41.143570 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.994610E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108580288 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59127. MP_Allreduce 11005 1515. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.045 297.196 297.197 qs_mol_dyn_low 1 2.0 0.003 0.003 296.612 296.629 qs_forces 11 3.9 0.005 0.006 296.518 296.521 qs_energies 11 4.9 0.001 0.001 287.127 287.144 scf_env_do_scf 11 5.9 0.001 0.001 264.967 264.984 velocity_verlet 10 3.0 0.001 0.001 213.153 213.162 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 139.933 139.935 init_scf_loop 11 6.9 0.000 0.000 124.770 124.773 prepare_preconditioner 11 7.9 0.000 0.000 119.719 119.739 make_preconditioner 11 8.9 0.000 0.000 119.719 119.739 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.646 116.893 qs_scf_new_mos 116 7.6 0.001 0.001 89.944 90.072 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.943 90.072 ot_scf_mini 116 9.6 0.003 0.004 85.201 85.261 dbcsr_multiply_generic 2485 12.5 0.209 0.222 82.789 83.330 cp_fm_upper_to_full 104 14.8 53.055 76.147 53.055 76.147 multiply_cannon 2485 13.5 0.692 0.724 58.569 58.925 multiply_cannon_loop 2485 14.5 0.475 0.517 54.920 56.221 ot_mini 116 10.6 0.001 0.001 44.143 44.204 dbcsr_complete_redistribute 393 12.7 3.990 4.024 29.671 42.832 copy_fm_to_dbcsr 208 11.6 0.002 0.002 26.321 39.409 rebuild_ks_matrix 127 8.3 0.001 0.001 37.902 37.937 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 37.901 37.937 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.027 36.849 mp_alltoall_i22 712 14.1 21.871 34.974 21.871 34.974 qs_ks_update_qs_env 127 7.6 0.001 0.001 34.859 34.889 cp_fm_cholesky_invert 11 10.9 33.418 33.424 33.418 33.424 mp_waitall_1 102768 16.8 28.375 32.365 28.375 32.365 qs_ot_get_p 127 10.4 0.001 0.001 25.622 25.687 qs_ot_get_derivative 116 11.6 0.002 0.002 23.986 24.043 qs_ot_p2m_diag 82 11.4 0.868 0.874 21.577 21.607 make_m2s 4970 13.5 0.076 0.078 19.583 20.548 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.218 20.231 calculate_rho_elec 127 8.7 0.476 0.476 20.217 20.231 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 19.114 20.185 ot_diis_step 116 11.6 0.022 0.022 20.125 20.125 make_images 4970 14.5 3.696 3.822 19.106 20.075 cp_dbcsr_syevd 82 12.4 0.005 0.006 19.863 19.866 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.363 19.550 apply_single 127 13.6 0.001 0.001 19.363 19.550 sum_up_and_integrate 127 10.3 0.318 0.320 19.395 19.491 integrate_v_rspace 127 11.3 0.004 0.004 19.076 19.174 multiply_cannon_multrec 9940 15.5 10.368 12.083 17.741 17.879 cp_fm_diag_elpa 82 13.4 0.001 0.001 16.691 16.691 cp_fm_diag_elpa_base 82 14.4 12.337 13.914 16.687 16.687 multiply_cannon_sync_h2d 9940 15.5 15.527 15.544 15.527 15.544 make_images_data 4970 15.5 0.060 0.065 10.477 12.309 hybrid_alltoall_any 5155 16.4 1.286 3.002 10.534 12.259 init_scf_run 11 5.9 0.000 0.001 12.073 12.073 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.073 12.073 pw_transfer 1535 11.6 0.092 0.093 11.346 11.352 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 11.114 11.121 fft_wrap_pw1pw2_140 519 13.2 3.054 3.088 9.865 9.880 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.325 9.368 mp_alltoall_d11v 2401 14.1 8.218 9.069 8.218 9.069 cp_fm_cholesky_decompose 22 10.9 8.929 8.959 8.929 8.959 wfi_extrapolate 11 7.9 0.001 0.001 8.952 8.952 dbcsr_mm_accdrv_process 20590 16.0 3.748 5.691 7.135 8.931 grid_integrate_task_list 127 12.3 8.504 8.687 8.504 8.687 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 8.091 8.092 density_rs2pw 127 9.7 0.005 0.005 7.876 8.013 calculate_dm_sparse 127 9.5 0.001 0.001 6.604 6.703 grid_collocate_task_list 127 9.7 6.307 6.341 6.307 6.341 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.214 6.289 fft3d_ps 1281 14.7 2.821 2.826 6.222 6.242 rs_scatter_matrices 138 9.7 3.624 4.701 6.015 6.239 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.004 6.072 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=297.197000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2698.454545, yerr=160.055879 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 1.262309E+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 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.048 88.321 88.334 qs_energies 1 2.0 0.000 0.000 87.901 87.915 ls_scf 1 3.0 0.000 0.000 86.900 86.914 dbcsr_multiply_generic 111 6.7 0.014 0.015 75.967 76.123 multiply_cannon 111 7.7 0.017 0.020 59.279 60.223 multiply_cannon_loop 111 8.7 0.210 0.226 55.859 57.019 ls_scf_main 1 4.0 0.000 0.000 52.340 52.340 density_matrix_trs4 2 5.0 0.002 0.003 46.860 46.934 ls_scf_init_scf 1 4.0 0.000 0.000 31.518 31.519 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.438 30.484 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.352 28.366 mp_waitall_1 11031 10.9 22.754 27.455 22.754 27.455 multiply_cannon_multrec 2664 9.7 8.198 8.952 15.488 17.165 multiply_cannon_sync_h2d 2664 9.7 13.606 15.185 13.606 15.185 make_m2s 222 7.7 0.008 0.011 13.035 13.513 make_images 222 8.7 0.098 0.108 13.013 13.495 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.712 11.986 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.752 9.745 make_images_data 222 9.7 0.004 0.005 7.627 8.269 dbcsr_mm_accdrv_process 4760 10.4 0.511 0.626 6.907 7.870 hybrid_alltoall_any 227 10.6 0.215 1.830 6.543 7.735 dbcsr_mm_accdrv_process_sort 4760 11.4 6.197 7.084 6.197 7.084 calculate_norms 4752 9.8 5.513 6.144 5.513 6.144 apply_matrix_preconditioner 6 5.3 0.000 0.003 5.006 5.118 mp_sum_l 807 5.4 3.114 4.684 3.114 4.684 compute_matrix_preconditioner 1 6.0 0.000 0.000 4.438 4.442 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.037 3.844 mp_irecv_dv 6231 10.9 2.020 3.829 2.020 3.829 acc_transpose_blocks 2664 9.7 0.019 0.021 3.460 3.766 acc_transpose_blocks_kernels 2664 10.7 0.037 0.041 3.350 3.653 make_images_sizes 222 9.7 0.000 0.000 0.794 3.644 mp_alltoall_i44 222 10.7 0.793 3.643 0.793 3.643 jit_kernel_transpose 1 13.0 3.313 3.614 3.313 3.614 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.318 3.453 arnoldi_extremal 4 6.8 0.000 0.000 3.293 3.313 arnoldi_normal_ev 4 7.8 0.001 0.003 3.293 3.312 build_subspace 16 8.4 0.009 0.012 3.193 3.196 ls_scf_post 1 4.0 0.000 0.000 3.041 3.055 ls_scf_store_result 1 5.0 0.000 0.000 2.861 2.900 dbcsr_special_finalize 555 9.7 0.005 0.006 2.264 2.791 dbcsr_merge_single_wm 555 10.7 0.451 0.575 2.255 2.783 make_images_pack 222 9.7 2.206 2.621 2.208 2.623 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.390 2.582 dbcsr_sort_data 658 11.4 2.063 2.490 2.063 2.490 dbcsr_matrix_vector_mult_local 304 10.0 2.078 2.459 2.080 2.461 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.262 2.328 buffer_matrices_ensure_size 222 8.7 1.751 2.044 1.751 2.044 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=88.334000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 2.105246E+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 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.074 94.460 94.461 qs_energies 1 2.0 0.000 0.000 93.888 93.898 ls_scf 1 3.0 0.000 0.001 92.543 92.554 dbcsr_multiply_generic 111 6.7 0.015 0.018 78.526 78.925 multiply_cannon 111 7.7 0.028 0.043 56.106 60.403 multiply_cannon_loop 111 8.7 0.116 0.122 52.837 56.301 ls_scf_main 1 4.0 0.000 0.001 55.612 55.618 density_matrix_trs4 2 5.0 0.002 0.003 49.906 50.085 ls_scf_init_scf 1 4.0 0.000 0.002 33.372 33.373 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.158 32.274 mp_waitall_1 9105 10.9 21.849 30.631 21.849 30.631 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.757 29.773 multiply_cannon_multrec 1332 9.7 13.117 16.276 22.300 26.603 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 12.077 20.799 make_m2s 222 7.7 0.006 0.007 15.538 16.244 make_images 222 8.7 1.577 1.966 15.508 16.216 dbcsr_mm_accdrv_process 4041 10.4 0.276 0.448 8.781 10.330 make_images_data 222 9.7 0.004 0.004 9.011 10.099 dbcsr_mm_accdrv_process_sort 4041 11.4 8.373 9.882 8.373 9.882 hybrid_alltoall_any 227 10.6 0.521 2.469 8.374 9.522 mp_sum_l 807 5.4 5.613 9.511 5.613 9.511 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.217 7.703 mp_irecv_dv 3311 11.0 3.197 7.648 3.197 7.648 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.322 7.509 calculate_norms 2376 9.8 5.996 6.695 5.996 6.695 multiply_cannon_sync_h2d 1332 9.7 4.801 6.117 4.801 6.117 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.033 5.256 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.391 4.747 arnoldi_extremal 4 6.8 0.000 0.000 4.651 4.669 arnoldi_normal_ev 4 7.8 0.001 0.004 4.651 4.669 build_subspace 16 8.4 0.014 0.021 4.398 4.400 ls_scf_post 1 4.0 0.001 0.007 3.560 3.570 compute_matrix_preconditioner 1 6.0 0.000 0.001 3.556 3.561 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.169 3.401 ls_scf_store_result 1 5.0 0.000 0.000 3.254 3.385 dbcsr_matrix_vector_mult_local 304 10.0 2.781 3.255 2.783 3.256 acc_transpose_blocks 1332 9.7 0.007 0.007 2.686 3.052 acc_transpose_blocks_kernels 1332 10.7 0.018 0.019 2.639 3.003 jit_kernel_transpose 1 13.0 2.621 2.985 2.621 2.985 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.592 2.697 mp_allgather_i34 111 8.7 0.971 2.484 0.971 2.484 make_images_pack 222 9.7 2.020 2.362 2.022 2.364 dbcsr_sort_data 436 11.2 1.836 2.105 1.836 2.105 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=94.461000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1751.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 2.790736E+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 3058 11181. 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 0.030 0.047 94.743 94.744 qs_energies 1 2.0 0.000 0.000 94.170 94.174 ls_scf 1 3.0 0.000 0.001 92.758 92.764 dbcsr_multiply_generic 111 6.7 0.016 0.018 77.458 77.735 ls_scf_main 1 4.0 0.000 0.001 58.203 58.208 multiply_cannon 111 7.7 0.034 0.073 53.372 57.779 multiply_cannon_loop 111 8.7 0.101 0.121 49.815 53.477 density_matrix_trs4 2 5.0 0.002 0.003 52.220 52.414 mp_waitall_1 7281 11.0 24.376 34.212 24.376 34.212 ls_scf_init_scf 1 4.0 0.000 0.001 30.967 30.970 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.798 29.862 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.410 27.425 multiply_cannon_multrec 888 9.7 12.534 15.272 21.024 24.386 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.569 23.094 make_m2s 222 7.7 0.006 0.007 17.361 18.561 make_images 222 8.7 1.973 2.312 17.323 18.523 hybrid_alltoall_any 227 10.6 0.622 2.868 9.575 11.150 make_images_data 222 9.7 0.004 0.004 9.981 10.994 dbcsr_mm_accdrv_process 3754 10.4 0.235 0.413 8.026 9.246 mp_sum_l 807 5.4 5.147 9.190 5.147 9.190 dbcsr_mm_accdrv_process_sort 3754 11.4 7.655 8.832 7.655 8.832 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.451 7.255 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.488 7.096 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.923 7.065 mp_irecv_dv 2335 11.1 2.473 7.051 2.473 7.051 multiply_cannon_sync_h2d 888 9.7 6.033 6.929 6.033 6.929 arnoldi_extremal 4 6.8 0.000 0.000 5.130 5.150 arnoldi_normal_ev 4 7.8 0.001 0.005 5.130 5.150 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.926 5.134 build_subspace 16 8.4 0.014 0.020 4.826 4.832 calculate_norms 1584 9.8 4.296 4.709 4.296 4.709 mp_allgather_i34 111 8.7 1.409 3.955 1.409 3.955 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.490 3.814 dbcsr_matrix_vector_mult_local 304 10.0 3.068 3.639 3.070 3.641 ls_scf_post 1 4.0 0.000 0.000 3.588 3.595 ls_scf_store_result 1 5.0 0.000 0.000 3.297 3.385 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.826 2.931 make_images_sizes 222 9.7 0.000 0.000 1.052 2.185 mp_alltoall_i44 222 10.7 1.052 2.185 1.052 2.185 dbcsr_sort_data 325 11.1 1.897 2.181 1.897 2.181 make_images_pack 222 9.7 1.811 2.093 1.814 2.096 dbcsr_data_release 9322 10.9 1.302 1.961 1.302 1.961 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.907 1.910 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.744000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2170.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 3.344880E+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 3058 13371. 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 0.032 0.074 97.353 97.355 qs_energies 1 2.0 0.000 0.000 96.736 96.741 ls_scf 1 3.0 0.000 0.001 95.084 95.091 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.866 79.137 ls_scf_main 1 4.0 0.000 0.001 59.264 59.266 multiply_cannon 111 7.7 0.068 0.158 51.661 56.253 density_matrix_trs4 2 5.0 0.002 0.003 53.204 53.337 multiply_cannon_loop 111 8.7 0.115 0.126 46.599 50.135 ls_scf_init_scf 1 4.0 0.000 0.002 32.568 32.569 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.341 31.429 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.890 28.902 mp_waitall_1 6369 11.0 23.005 28.814 23.005 28.814 multiply_cannon_multrec 1332 9.7 14.132 17.279 21.950 24.664 make_m2s 222 7.7 0.006 0.008 21.332 22.717 make_images 222 8.7 3.131 3.603 21.282 22.669 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.282 17.005 make_images_data 222 9.7 0.004 0.004 11.924 13.482 hybrid_alltoall_any 227 10.6 0.797 3.748 11.289 13.143 dbcsr_mm_accdrv_process 3641 10.4 0.212 0.410 7.460 8.970 dbcsr_mm_accdrv_process_sort 3641 11.4 7.086 8.551 7.086 8.551 mp_sum_l 807 5.4 4.283 8.264 4.283 8.264 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.288 6.670 multiply_cannon_sync_h2d 1332 9.7 5.498 6.049 5.498 6.049 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.083 5.990 mp_irecv_dv 3229 10.9 2.061 5.907 2.061 5.907 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.630 5.539 arnoldi_extremal 4 6.8 0.000 0.000 5.289 5.302 arnoldi_normal_ev 4 7.8 0.001 0.005 5.289 5.302 build_subspace 16 8.4 0.014 0.021 4.948 4.955 apply_matrix_preconditioner 6 5.3 0.000 0.001 4.521 4.718 mp_allgather_i34 111 8.7 2.178 4.612 2.178 4.612 calculate_norms 2376 9.8 4.197 4.531 4.197 4.531 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.660 3.956 dbcsr_matrix_vector_mult_local 304 10.0 3.278 3.759 3.280 3.760 dbcsr_sort_data 658 11.4 3.130 3.441 3.130 3.441 ls_scf_post 1 4.0 0.000 0.000 3.252 3.257 dbcsr_special_finalize 555 9.7 0.006 0.007 2.871 3.246 dbcsr_merge_single_wm 555 10.7 0.538 0.668 2.862 3.238 ls_scf_store_result 1 5.0 0.000 0.000 3.002 3.058 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.983 3.025 dbcsr_data_release 10477 10.7 1.575 2.393 1.575 2.393 dbcsr_finalize 304 7.8 0.049 0.061 1.805 1.963 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.355000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2742.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 4.678750E+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 3049 15663. 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 0.039 0.057 99.786 99.788 qs_energies 1 2.0 0.000 0.000 98.981 98.986 ls_scf 1 3.0 0.000 0.000 97.029 97.039 dbcsr_multiply_generic 111 6.7 0.018 0.019 78.489 78.728 ls_scf_main 1 4.0 0.000 0.000 62.591 62.592 multiply_cannon 111 7.7 0.096 0.177 55.589 60.190 density_matrix_trs4 2 5.0 0.002 0.003 55.485 55.573 multiply_cannon_loop 111 8.7 0.070 0.079 51.170 53.442 mp_waitall_1 5436 11.0 27.092 32.142 27.092 32.142 ls_scf_init_scf 1 4.0 0.000 0.000 30.811 30.816 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.578 29.619 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.417 27.427 multiply_cannon_multrec 444 9.7 14.132 16.242 21.146 23.516 make_m2s 222 7.7 0.004 0.005 17.884 20.291 make_images 222 8.7 3.716 4.422 17.822 20.231 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.661 16.502 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.987 15.101 make_images_data 222 9.7 0.003 0.004 10.244 12.461 hybrid_alltoall_any 227 10.6 0.787 3.783 9.994 12.358 multiply_cannon_sync_h2d 444 9.7 6.571 8.245 6.571 8.245 dbcsr_mm_accdrv_process 3003 10.4 0.187 0.386 6.722 7.833 dbcsr_mm_accdrv_process_sort 3003 11.4 6.396 7.486 6.396 7.486 mp_allgather_i34 111 8.7 2.656 6.998 2.656 6.998 arnoldi_extremal 4 6.8 0.000 0.000 5.934 5.944 arnoldi_normal_ev 4 7.8 0.003 0.006 5.934 5.944 build_subspace 16 8.4 0.015 0.020 5.538 5.545 mp_sum_l 807 5.4 3.068 4.805 3.068 4.805 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.540 4.703 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.279 4.479 dbcsr_matrix_vector_mult_local 304 10.0 3.837 4.278 3.839 4.280 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.620 3.846 mp_irecv_dv 1241 11.2 1.603 3.825 1.603 3.825 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.156 3.743 calculate_norms 792 9.8 3.531 3.670 3.531 3.670 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.552 3.649 ls_scf_post 1 4.0 0.000 0.000 3.627 3.631 make_images_sizes 222 9.7 0.000 0.000 1.004 3.587 mp_alltoall_i44 222 10.7 1.004 3.586 1.004 3.586 ls_scf_store_result 1 5.0 0.000 0.000 3.408 3.453 dbcsr_finalize 304 7.8 0.062 0.078 2.191 2.278 dbcsr_merge_all 275 8.9 0.473 0.525 2.042 2.116 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.055 2.056 rebuild_ks_matrix 3 7.3 0.000 0.000 2.023 2.024 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.023 2.024 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.788000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3733.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 8.708600E+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 3043 21950. 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 0.076 0.099 109.302 109.304 qs_energies 1 2.0 0.000 0.000 107.865 107.878 ls_scf 1 3.0 0.000 0.000 104.962 104.974 dbcsr_multiply_generic 111 6.7 0.023 0.028 78.374 78.483 ls_scf_main 1 4.0 0.000 0.000 66.238 66.239 density_matrix_trs4 2 5.0 0.002 0.003 57.069 57.121 multiply_cannon 111 7.7 0.137 0.242 50.876 52.739 multiply_cannon_loop 111 8.7 0.070 0.083 47.427 48.396 ls_scf_init_scf 1 4.0 0.000 0.000 35.069 35.070 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.501 33.513 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.684 30.700 mp_waitall_1 4527 11.1 22.194 25.885 22.194 25.885 make_m2s 222 7.7 0.005 0.005 23.848 24.818 make_images 222 8.7 4.587 4.957 23.742 24.709 multiply_cannon_multrec 444 9.7 17.889 18.607 22.477 23.087 hybrid_alltoall_any 227 10.6 1.659 3.629 12.891 15.529 make_images_data 222 9.7 0.003 0.004 13.082 15.491 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.537 12.166 multiply_cannon_sync_h2d 444 9.7 8.848 8.886 8.848 8.886 arnoldi_extremal 4 6.8 0.000 0.000 7.655 7.676 arnoldi_normal_ev 4 7.8 0.003 0.009 7.654 7.675 build_subspace 16 8.4 0.026 0.036 7.098 7.109 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.758 5.883 dbcsr_matrix_vector_mult_local 304 10.0 5.301 5.598 5.304 5.600 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.011 5.258 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.125 5.213 dbcsr_mm_accdrv_process 1814 10.4 0.256 0.322 4.422 4.549 dbcsr_mm_accdrv_process_sort 1814 11.4 4.120 4.243 4.120 4.243 ls_scf_post 1 4.0 0.000 0.000 3.655 3.667 mp_allgather_i34 111 8.7 1.100 3.492 1.100 3.492 make_images_sizes 222 9.7 0.000 0.000 1.432 3.431 mp_alltoall_i44 222 10.7 1.432 3.431 1.432 3.431 ls_scf_store_result 1 5.0 0.000 0.000 3.403 3.416 calculate_norms 792 9.8 3.240 3.281 3.240 3.281 dbcsr_finalize 304 7.8 0.082 0.090 3.084 3.158 dbcsr_merge_all 275 8.9 0.892 0.911 2.869 2.939 dbcsr_complete_redistribute 5 7.6 1.426 1.455 2.756 2.875 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.873 2.873 dbcsr_data_release 12724 10.6 2.334 2.866 2.334 2.866 matrix_ls_to_qs 2 6.0 0.000 0.000 2.397 2.519 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.501 2.502 dbcsr_sort_data 325 11.1 2.445 2.500 2.445 2.500 rebuild_ks_matrix 3 7.3 0.000 0.000 2.433 2.434 qs_ks_build_kohn_sham_matrix 3 8.3 0.021 0.024 2.433 2.434 dbcsr_new_transposed 4 7.5 0.243 0.251 2.265 2.276 dbcsr_frobenius_norm 74 6.6 2.057 2.135 2.194 2.233 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.201 dbcsr_add_anytype 103 7.2 0.860 0.891 2.132 2.200 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.304000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6868.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b57c5e43b9bf4ebed0002f204f6eebfbc3082088_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 590.340096E+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 13152 2094. MP_Sync 1064 MP_Alltoall 2588 926978254. 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 0.026 0.090 225.302 225.304 qs_mol_dyn_low 1 2.0 0.003 0.013 224.411 224.434 qs_forces 5 3.8 0.005 0.043 224.319 224.329 qs_energies 5 4.8 0.001 0.055 221.366 221.421 scf_env_do_scf 5 5.8 0.000 0.001 207.065 207.068 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 180.405 180.407 qs_scf_new_mos 105 7.6 0.000 0.001 139.875 140.065 qs_scf_loop_do_ot 105 8.6 0.001 0.001 139.874 140.065 dbcsr_multiply_generic 1445 12.2 0.131 0.139 132.122 132.570 ot_scf_mini 105 9.6 0.003 0.003 130.136 130.289 multiply_cannon 1445 13.2 0.273 0.284 113.011 115.266 multiply_cannon_loop 1445 14.2 2.408 2.537 111.349 112.490 velocity_verlet 4 3.0 0.001 0.001 105.846 105.847 ot_mini 105 10.6 0.001 0.002 59.745 59.849 multiply_cannon_multrec 69360 15.2 31.763 37.484 39.415 44.872 mp_waitall_1 488190 16.1 34.057 42.717 34.057 42.717 qs_ot_get_p 112 10.4 0.001 0.001 40.827 41.172 qs_ot_get_derivative 55 11.6 0.001 0.001 38.226 38.341 multiply_cannon_sync_h2d 69360 15.2 30.720 35.290 30.720 35.290 multiply_cannon_metrocomm3 69360 15.2 0.191 0.201 24.958 34.169 qs_ot_p2m_diag 40 11.0 0.020 0.031 29.959 30.033 rebuild_ks_matrix 110 8.4 0.000 0.000 29.003 29.183 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.015 29.003 29.183 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.640 26.799 cp_dbcsr_syevd 40 12.0 0.002 0.002 26.720 26.721 init_scf_loop 7 6.6 0.000 0.001 26.630 26.631 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 22.839 23.157 apply_single 62 13.6 0.000 0.000 22.839 23.157 prepare_preconditioner 7 7.6 0.000 0.000 21.737 21.766 make_preconditioner 7 8.6 0.000 0.000 21.737 21.766 cp_fm_syevd 40 13.0 0.000 0.001 21.254 21.393 ot_new_cg_direction 55 11.6 0.001 0.002 20.828 20.828 qs_rho_update_rho_low 110 7.6 0.001 0.002 17.307 17.619 calculate_rho_elec 110 8.6 0.030 0.033 17.307 17.619 cp_fm_redistribute_end 40 14.0 8.326 16.614 8.330 16.615 cp_fm_syevd_base 40 14.0 8.278 16.567 8.278 16.567 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.506 14.731 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.533 14.611 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.632 13.706 mp_sum_l 4684 12.4 12.425 13.200 12.425 13.200 rs_pw_transfer 690 11.5 0.011 0.013 11.541 12.907 density_rs2pw 110 9.6 0.006 0.007 11.434 12.838 pw_transfer 1645 12.4 0.083 0.101 11.953 12.124 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 11.810 11.985 init_scf_run 5 5.8 0.000 0.001 11.613 11.613 scf_env_initial_rho_setup 5 6.8 0.002 0.004 11.612 11.613 calculate_dm_sparse 110 9.5 0.000 0.001 11.265 11.406 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.543 10.629 qs_vxc_create 110 10.4 0.002 0.003 10.382 10.416 cp_fm_cholesky_invert 7 10.6 10.244 10.251 10.244 10.251 fft_wrap_pw1pw2_240 915 15.0 1.203 1.280 9.993 10.193 check_diag 80 13.5 8.592 8.858 9.555 9.693 sum_up_and_integrate 60 10.3 0.028 0.031 8.691 8.701 integrate_v_rspace 60 11.3 0.002 0.002 8.663 8.674 dbcsr_mm_accdrv_process 154766 15.8 4.060 4.235 7.521 8.295 fft3d_pb 915 16.0 2.393 2.563 8.033 8.193 multiply_cannon_metrocomm1 69360 15.2 0.091 0.096 4.685 7.813 xc_rho_set_and_dset_create 110 12.4 0.077 0.099 7.260 7.507 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 7.281 7.295 acc_transpose_blocks 69360 15.2 0.359 0.376 6.371 7.159 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.922 6.973 xc_vxc_pw_create 60 11.3 0.039 0.049 6.931 6.965 make_full_single_inverse 7 9.6 0.001 0.001 6.929 6.958 make_m2s 2890 13.2 0.081 0.090 6.246 6.824 calculate_first_density_matrix 1 7.0 0.000 0.002 6.781 6.798 make_images 2890 14.2 0.241 0.262 6.137 6.715 xc_pw_derive 510 13.4 0.006 0.007 6.144 6.212 mp_waitany 7680 13.5 4.510 6.050 4.510 6.050 mp_alltoall_z22v 2340 17.7 5.074 5.402 5.074 5.402 acc_transpose_blocks_kernels 69360 16.2 0.857 0.910 4.458 5.156 potential_pw2rs 60 12.3 0.003 0.003 4.842 4.876 multiply_cannon_metrocomm4 67915 15.2 0.181 0.196 2.059 4.627 grid_collocate_task_list 110 9.6 4.092 4.598 4.092 4.598 wfi_extrapolate 5 7.8 0.000 0.002 4.509 4.509 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=225.304000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.400000, yerr=3.200000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: b57c5e43b9bf4ebed0002f204f6eebfbc3082088 Summary: empty Status: OK