=== 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: 97946bc07285c4fbfd3a5916190ff8dec9cb9ad5 ################# 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/01 job id: 45432050 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/02 job id: 45432052 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/03 job id: 45432053 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/04 job id: 45432054 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/05 job id: 45432056 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/06 job id: 45432058 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/07 job id: 45432060 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/08 job id: 45432061 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/09 job id: 45432062 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/10 job id: 45432063 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/11 job id: 45432064 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/12 job id: 45432065 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/13 job id: 45432066 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/14 job id: 45432067 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/15 job id: 45432068 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/16 job id: 45432069 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/17 job id: 45432070 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/18 job id: 45432071 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/19 job id: 45432072 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/20 job id: 45432077 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/21 job id: 45432078 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/22 job id: 45432079 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/23 job id: 45432080 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/24 job id: 45432083 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/25 job id: 45432084 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/26 job id: 45432085 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/27 job id: 45432086 --- 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/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 0.022 0.036 133.955 133.956 farming_run 1 2.0 133.405 133.410 133.924 133.928 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.460326E+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 0.025 115.267 115.267 qs_energies 1 2.0 0.000 0.000 115.073 115.075 mp2_main 1 3.0 0.000 0.000 112.983 112.985 mp2_gpw_main 1 4.0 0.023 0.028 111.958 111.960 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 93.292 93.418 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.412 55.536 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.167 41.711 46.684 get_2c_integrals 1 6.0 0.000 0.000 36.973 37.709 integrate_v_rspace 273 8.0 0.436 0.451 25.067 29.738 pw_transfer 6555 10.6 0.372 0.378 27.312 27.764 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.979 26.362 grid_integrate_task_list 273 9.0 20.895 25.997 20.895 25.997 fft_wrap_pw1pw2_100 2178 12.4 1.178 1.255 23.502 23.909 compute_2c_integrals 1 7.0 0.002 0.003 19.368 19.370 compute_2c_integrals_loop_lm 1 8.0 0.003 0.005 18.827 19.095 mp2_eri_2c_integrate_gpw 1 9.0 2.380 2.425 18.825 19.094 rpa_ri_compute_en 1 5.0 0.001 0.002 18.550 18.647 cp_fm_cholesky_decompose 12 8.2 17.594 18.369 17.594 18.369 cholesky_decomp 1 7.0 0.000 0.000 16.459 17.234 fft3d_s 5443 13.4 16.099 16.426 16.121 16.449 ao_to_mo_and_store_B_mult_1 272 7.0 10.866 15.589 10.866 15.589 calculate_wavefunction 272 8.0 5.475 5.551 12.573 13.217 rpa_num_int 1 6.0 0.000 0.001 10.581 10.591 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.555 10.579 calc_mat_Q 8 8.0 0.000 0.000 9.376 9.478 contract_S_to_Q 8 9.0 0.000 0.000 8.797 8.897 calc_potential_gpw 544 9.5 0.005 0.006 8.240 8.580 parallel_gemm_fm 14 9.1 0.000 0.000 8.372 8.466 parallel_gemm_fm_cosma 14 10.1 8.372 8.466 8.372 8.466 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.178 8.397 potential_pw2rs 545 10.0 0.107 0.110 7.641 8.257 create_integ_mat 1 6.0 0.014 0.027 7.748 7.757 collocate_single_gaussian 272 10.0 0.040 0.042 7.448 7.691 array2fm 1 7.0 0.000 0.000 6.693 7.178 pw_scatter_s 2720 13.7 4.439 4.570 4.439 4.570 pw_gather_s 2722 13.2 3.858 4.191 3.858 4.191 array2fm_buffer_send 1 8.0 2.958 3.169 2.958 3.169 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.958465, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2731.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 0.028 0.039 395.948 395.949 farming_run 1 2.0 395.285 395.289 395.914 395.916 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.223840E+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 14327492. 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 0.009 0.028 209.711 209.712 qs_energies 1 2.0 0.000 0.000 209.508 209.515 scf_env_do_scf 1 3.0 0.000 0.000 106.569 106.569 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.693 105.702 rebuild_ks_matrix 4 6.0 0.000 0.000 105.691 105.700 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.065 105.691 105.700 hfx_ks_matrix 4 8.0 0.001 0.001 105.310 105.315 integrate_four_center 4 9.0 0.144 0.471 105.309 105.315 mp2_main 1 3.0 0.000 0.000 102.649 102.657 mp2_gpw_main 1 4.0 0.032 0.046 101.808 101.816 integrate_four_center_main 4 10.0 0.100 0.517 96.514 99.516 integrate_four_center_bin 263 11.0 96.414 98.999 96.414 98.999 init_scf_loop 1 4.0 0.000 0.000 92.298 92.298 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 75.051 76.068 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.505 55.523 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.157 42.346 47.316 integrate_v_rspace 95 8.0 0.398 0.562 28.712 33.486 pw_transfer 2240 10.6 0.145 0.186 29.946 30.390 ao_to_mo_and_store_B_mult_1 91 7.0 10.478 29.581 10.478 29.581 fft_wrap_pw1pw2 1868 11.4 0.018 0.023 28.948 29.409 grid_integrate_task_list 95 9.0 23.999 28.990 23.999 28.990 mp2_ri_gpw_compute_en 1 5.0 0.054 0.063 26.612 28.340 fft_wrap_pw1pw2_100 730 12.4 1.283 1.493 26.668 27.152 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.853 1.906 24.953 24.962 get_2c_integrals 1 6.0 0.000 0.000 20.464 20.481 compute_2c_integrals 1 7.0 0.002 0.003 19.449 19.451 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.885 19.326 mp2_eri_2c_integrate_gpw 1 9.0 1.733 1.899 18.884 19.324 fft3d_s 1823 13.4 18.438 18.872 18.452 18.885 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.269 14.269 calculate_wavefunction 91 8.0 2.024 2.073 9.752 9.985 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.592 8.767 9.213 potential_pw2rs 186 10.0 0.034 0.036 8.652 9.213 local_gemm 172 8.0 8.212 8.633 8.212 8.633 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.260 8.586 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.526 7.937 8.387 calc_potential_gpw 182 9.5 0.002 0.003 7.915 8.208 collocate_single_gaussian 91 10.0 0.017 0.021 7.896 8.153 mp_sync 37 10.5 3.449 6.451 3.449 6.451 mp2_ri_gpw_compute_en_ener 172 7.0 6.350 6.445 6.350 6.445 mp_sendrecv_dm3 2068 8.0 5.967 6.438 5.967 6.438 pw_gather_s 912 13.2 4.923 5.345 4.923 5.345 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.800844, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.128768E+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 0.015 0.031 49.857 49.859 qs_mol_dyn_low 1 2.0 0.004 0.009 49.571 49.578 qs_forces 11 3.9 0.003 0.008 49.252 49.253 qs_energies 11 4.9 0.002 0.011 47.803 47.819 scf_env_do_scf 11 5.9 0.000 0.001 41.702 41.702 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 39.683 39.683 dbcsr_multiply_generic 2286 12.5 0.109 0.116 30.252 30.664 qs_scf_new_mos 108 7.5 0.000 0.001 29.784 30.050 qs_scf_loop_do_ot 108 8.5 0.000 0.001 29.783 30.049 ot_scf_mini 108 9.5 0.002 0.004 28.217 28.404 multiply_cannon 2286 13.5 0.195 0.207 24.753 26.173 multiply_cannon_loop 2286 14.5 1.454 1.558 24.084 25.522 velocity_verlet 10 3.0 0.001 0.002 24.229 24.232 ot_mini 108 10.5 0.001 0.002 16.585 16.840 qs_ot_get_derivative 108 11.5 0.001 0.004 13.798 13.966 mp_waitall_1 245248 16.5 6.818 12.891 6.818 12.891 multiply_cannon_metrocomm3 54864 15.5 0.071 0.077 5.041 11.684 multiply_cannon_multrec 54864 15.5 4.415 6.873 7.796 11.333 rebuild_ks_matrix 119 8.3 0.000 0.000 7.826 7.971 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.014 7.825 7.971 qs_ot_get_p 119 10.4 0.001 0.001 7.405 7.680 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.905 7.042 multiply_cannon_sync_h2d 54864 15.5 6.181 6.790 6.181 6.790 mp_sum_l 7207 12.9 3.610 5.287 3.610 5.287 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.804 5.230 qs_ot_p2m_diag 50 11.0 0.004 0.006 4.949 4.981 init_scf_run 11 5.9 0.000 0.001 4.842 4.842 scf_env_initial_rho_setup 11 6.9 0.001 0.003 4.841 4.842 dbcsr_mm_accdrv_process 76910 16.1 1.078 1.808 3.301 4.687 sum_up_and_integrate 119 10.3 0.012 0.015 4.494 4.503 integrate_v_rspace 119 11.3 0.002 0.003 4.482 4.490 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.317 4.437 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.315 4.315 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.125 4.126 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.032 4.114 calculate_rho_elec 119 8.7 0.012 0.017 4.031 4.113 cp_fm_redistribute_end 50 14.0 2.107 4.101 2.113 4.104 cp_fm_diag_elpa_base 50 14.0 1.986 3.995 1.990 4.003 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.846 3.076 apply_single 119 13.6 0.000 0.000 2.846 3.076 calculate_dm_sparse 119 9.5 0.000 0.001 2.885 3.024 jit_kernel_multiply 13 15.8 2.166 2.871 2.166 2.871 rs_pw_transfer 974 11.9 0.011 0.012 2.672 2.767 calculate_first_density_matrix 1 7.0 0.002 0.015 2.749 2.754 ot_diis_step 108 11.5 0.006 0.009 2.596 2.597 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.344 2.346 acc_transpose_blocks 54864 15.5 0.219 0.254 1.731 2.279 density_rs2pw 119 9.7 0.004 0.005 2.140 2.272 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.123 2.169 grid_integrate_task_list 119 12.3 2.047 2.157 2.047 2.157 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.071 2.131 wfi_extrapolate 11 7.9 0.001 0.003 2.032 2.032 init_scf_loop 11 6.9 0.000 0.001 2.002 2.002 multiply_cannon_metrocomm1 54864 15.5 0.054 0.059 1.153 1.979 potential_pw2rs 119 12.3 0.004 0.004 1.825 1.835 mp_sum_d 4135 12.0 1.142 1.778 1.142 1.778 pw_transfer 1439 11.6 0.051 0.056 1.691 1.762 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.616 1.688 make_m2s 4572 13.5 0.053 0.056 1.572 1.617 make_images 4572 14.5 0.133 0.139 1.491 1.534 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.476 1.497 mp_waitany 12084 13.8 1.253 1.429 1.253 1.429 grid_collocate_task_list 119 9.7 1.289 1.353 1.289 1.353 mp_alltoall_d11v 2130 13.8 1.156 1.329 1.156 1.329 fft3d_ps 1201 14.6 0.371 0.479 1.259 1.327 fft_wrap_pw1pw2_140 487 13.2 0.184 0.200 1.244 1.316 acc_transpose_blocks_kernels 54864 16.5 0.236 0.379 0.813 1.127 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.618 1.035 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=49.859000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 487.190528E+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 1030423. 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 0.019 0.035 39.480 39.481 qs_mol_dyn_low 1 2.0 0.003 0.005 39.162 39.168 qs_forces 11 3.9 0.002 0.004 39.060 39.062 qs_energies 11 4.9 0.002 0.008 37.343 37.346 scf_env_do_scf 11 5.9 0.001 0.001 32.018 32.019 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 29.531 29.533 dbcsr_multiply_generic 2286 12.5 0.099 0.105 21.513 21.871 qs_scf_new_mos 108 7.5 0.001 0.001 20.218 20.463 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.217 20.462 ot_scf_mini 108 9.5 0.003 0.003 19.313 19.484 velocity_verlet 10 3.0 0.001 0.001 18.329 18.330 multiply_cannon 2286 13.5 0.209 0.217 16.471 18.128 multiply_cannon_loop 2286 14.5 0.897 0.972 15.367 16.861 ot_mini 108 10.5 0.001 0.002 11.835 12.073 mp_waitall_1 200699 16.5 5.738 10.877 5.738 10.877 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.216 9.616 qs_ot_get_derivative 108 11.5 0.001 0.001 9.374 9.550 multiply_cannon_multrec 27432 15.5 1.970 4.287 5.848 8.808 rebuild_ks_matrix 119 8.3 0.000 0.000 7.270 7.404 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.017 7.269 7.404 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.404 6.527 dbcsr_mm_accdrv_process 47894 16.0 3.026 5.205 3.806 5.637 qs_ot_get_p 119 10.4 0.001 0.001 4.593 4.834 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.603 4.452 sum_up_and_integrate 119 10.3 0.024 0.029 4.256 4.265 integrate_v_rspace 119 11.3 0.002 0.002 4.232 4.244 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.046 4.114 apply_single 119 13.6 0.000 0.000 3.046 4.114 init_scf_run 11 5.9 0.000 0.002 4.030 4.030 scf_env_initial_rho_setup 11 6.9 0.001 0.003 4.030 4.030 mp_sum_l 7207 12.9 2.113 4.015 2.113 4.015 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.889 3.916 calculate_rho_elec 119 8.7 0.021 0.024 3.888 3.915 rs_pw_transfer 974 11.9 0.010 0.011 2.711 3.228 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.104 3.122 multiply_cannon_sync_h2d 27432 15.5 2.194 2.887 2.194 2.887 density_rs2pw 119 9.7 0.004 0.005 2.209 2.731 make_m2s 4572 13.5 0.052 0.054 2.487 2.711 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.667 2.667 make_images 4572 14.5 0.200 0.236 2.399 2.620 calculate_first_density_matrix 1 7.0 0.000 0.002 2.518 2.521 init_scf_loop 11 6.9 0.001 0.004 2.466 2.471 ot_diis_step 108 11.5 0.011 0.016 2.409 2.409 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.333 2.333 cp_fm_redistribute_end 50 14.0 1.180 2.302 1.184 2.305 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.191 2.281 calculate_dm_sparse 119 9.5 0.000 0.001 2.169 2.248 cp_fm_diag_elpa_base 50 14.0 1.087 2.200 1.117 2.246 jit_kernel_multiply 11 16.2 0.728 2.067 0.728 2.067 pw_transfer 1439 11.6 0.065 0.069 2.006 2.044 potential_pw2rs 119 12.3 0.006 0.006 1.977 1.989 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.914 1.954 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.950 1.952 grid_integrate_task_list 119 12.3 1.842 1.944 1.842 1.944 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.725 1.767 acc_transpose_blocks 27432 15.5 0.107 0.111 1.314 1.639 prepare_preconditioner 11 7.9 0.000 0.000 1.541 1.568 make_preconditioner 11 8.9 0.000 0.001 1.541 1.568 make_images_data 4572 15.5 0.045 0.052 1.146 1.559 fft_wrap_pw1pw2_140 487 13.2 0.202 0.215 1.499 1.537 fft3d_ps 1201 14.6 0.525 0.581 1.465 1.499 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.438 1.499 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.000 1.485 wfi_extrapolate 11 7.9 0.001 0.001 1.448 1.448 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.375 1.381 grid_collocate_task_list 119 9.7 1.229 1.369 1.229 1.369 mp_alltoall_d11v 2130 13.8 1.194 1.324 1.194 1.324 mp_allgather_i34 2286 14.5 0.547 1.292 0.547 1.292 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.177 1.226 mp_waitany 5720 13.7 0.562 1.110 0.562 1.110 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.151 0.578 1.087 mp_sum_d 4135 12.0 0.619 1.082 0.619 1.082 acc_transpose_blocks_kernels 27432 16.5 0.181 0.271 0.785 1.035 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.024 1.025 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.926 0.940 mp_alltoall_z22v 1201 16.6 0.730 0.820 0.730 0.820 rs_pw_transfer_PW2RS_50 119 14.3 0.591 0.611 0.750 0.791 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.481000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.727273, yerr=1.135454 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 525.336576E+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 922641. 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 0.052 0.087 32.380 32.382 qs_mol_dyn_low 1 2.0 0.003 0.004 31.959 31.966 qs_forces 11 3.9 0.003 0.003 31.894 31.895 qs_energies 11 4.9 0.003 0.009 30.324 30.326 scf_env_do_scf 11 5.9 0.001 0.001 25.489 25.489 scf_env_do_scf_inner_loop 108 6.5 0.003 0.009 22.937 22.938 dbcsr_multiply_generic 2286 12.5 0.092 0.094 16.341 16.435 velocity_verlet 10 3.0 0.001 0.001 15.159 15.161 qs_scf_new_mos 108 7.5 0.001 0.001 14.737 14.755 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.736 14.755 ot_scf_mini 108 9.5 0.002 0.003 14.026 14.041 multiply_cannon 2286 13.5 0.195 0.203 13.146 13.892 multiply_cannon_loop 2286 14.5 0.630 0.658 12.375 13.164 ot_mini 108 10.5 0.001 0.001 8.626 8.642 qs_ot_get_derivative 108 11.5 0.001 0.001 7.143 7.157 multiply_cannon_multrec 18288 15.5 1.943 2.881 6.746 7.041 rebuild_ks_matrix 119 8.3 0.000 0.000 6.391 6.405 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.391 6.405 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.654 5.667 dbcsr_mm_accdrv_process 38222 16.0 3.992 5.244 4.717 5.562 sum_up_and_integrate 119 10.3 0.031 0.032 3.960 3.965 integrate_v_rspace 119 11.3 0.002 0.003 3.928 3.938 init_scf_run 11 5.9 0.000 0.002 3.633 3.633 scf_env_initial_rho_setup 11 6.9 0.002 0.003 3.632 3.633 mp_waitall_1 158411 16.6 2.576 3.487 2.576 3.487 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.464 3.478 calculate_rho_elec 119 8.7 0.031 0.032 3.463 3.477 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.679 3.323 qs_ot_get_p 119 10.4 0.001 0.001 3.245 3.268 rs_pw_transfer 974 11.9 0.009 0.010 2.285 2.546 init_scf_loop 11 6.9 0.002 0.011 2.535 2.537 calculate_first_density_matrix 1 7.0 0.002 0.018 2.406 2.417 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.045 2.369 apply_single 119 13.6 0.000 0.000 2.045 2.369 multiply_cannon_metrocomm3 18288 15.5 0.044 0.047 1.438 2.323 density_rs2pw 119 9.7 0.004 0.004 1.973 2.235 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.206 2.211 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.946 1.946 pw_transfer 1439 11.6 0.065 0.070 1.913 1.922 make_m2s 4572 13.5 0.044 0.045 1.743 1.895 grid_integrate_task_list 119 12.3 1.801 1.870 1.801 1.870 jit_kernel_multiply 10 16.2 0.675 1.857 0.675 1.857 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.820 1.832 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.825 1.826 calculate_dm_sparse 119 9.5 0.000 0.001 1.799 1.815 make_images 4572 14.5 0.189 0.201 1.658 1.809 potential_pw2rs 119 12.3 0.007 0.008 1.721 1.727 prepare_preconditioner 11 7.9 0.000 0.001 1.722 1.724 make_preconditioner 11 8.9 0.000 0.002 1.722 1.724 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.693 1.705 cp_fm_diag_elpa_base 50 14.0 1.670 1.681 1.691 1.704 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.677 1.682 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.575 1.661 multiply_cannon_sync_h2d 18288 15.5 1.421 1.581 1.421 1.581 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.524 1.532 fft_wrap_pw1pw2_140 487 13.2 0.254 0.260 1.463 1.475 ot_diis_step 108 11.5 0.011 0.011 1.463 1.463 mp_sum_l 7207 12.9 1.080 1.424 1.080 1.424 acc_transpose_blocks 18288 15.5 0.075 0.077 1.370 1.411 grid_collocate_task_list 119 9.7 1.212 1.349 1.212 1.349 fft3d_ps 1201 14.6 0.530 0.546 1.307 1.316 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.182 1.186 wfi_extrapolate 11 7.9 0.001 0.001 1.167 1.167 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.970 0.971 acc_transpose_blocks_kernels 18288 16.5 0.208 0.216 0.934 0.970 make_images_data 4572 15.5 0.044 0.048 0.786 0.953 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.862 0.885 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.676 0.863 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.810 0.812 mp_waitany 9880 13.7 0.529 0.792 0.529 0.792 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.128 0.518 0.775 jit_kernel_transpose 5 15.6 0.726 0.755 0.726 0.755 mp_alltoall_d11v 2130 13.8 0.660 0.742 0.660 0.742 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.659 0.734 cp_fm_cholesky_invert 11 10.9 0.711 0.715 0.711 0.715 mp_alltoall_z22v 1201 16.6 0.633 0.697 0.633 0.697 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.382000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.636364, yerr=2.226809 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 549.957632E+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 0.015 0.032 34.662 34.663 qs_mol_dyn_low 1 2.0 0.003 0.003 34.419 34.426 qs_forces 11 3.9 0.003 0.003 34.359 34.360 qs_energies 11 4.9 0.002 0.009 32.636 32.642 scf_env_do_scf 11 5.9 0.001 0.001 27.652 27.654 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 24.295 24.295 dbcsr_multiply_generic 2286 12.5 0.096 0.100 17.787 17.903 velocity_verlet 10 3.0 0.001 0.001 17.595 17.597 qs_scf_new_mos 108 7.5 0.001 0.001 15.881 15.936 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.880 15.936 ot_scf_mini 108 9.5 0.002 0.003 14.949 14.995 multiply_cannon 2286 13.5 0.229 0.277 14.345 14.712 multiply_cannon_loop 2286 14.5 0.937 0.970 13.408 13.782 ot_mini 108 10.5 0.001 0.001 9.012 9.070 multiply_cannon_multrec 27432 15.5 2.328 2.973 8.569 8.923 dbcsr_mm_accdrv_process 47916 15.9 5.151 6.819 6.145 7.350 qs_ot_get_derivative 108 11.5 0.001 0.001 7.222 7.268 rebuild_ks_matrix 119 8.3 0.000 0.000 6.562 6.616 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.561 6.616 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.822 5.869 sum_up_and_integrate 119 10.3 0.036 0.038 3.797 3.803 integrate_v_rspace 119 11.3 0.003 0.004 3.761 3.768 init_scf_run 11 5.9 0.000 0.001 3.625 3.625 scf_env_initial_rho_setup 11 6.9 0.001 0.003 3.624 3.625 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.509 3.541 calculate_rho_elec 119 8.7 0.040 0.046 3.508 3.541 qs_ot_get_p 119 10.4 0.001 0.001 3.341 3.407 init_scf_loop 11 6.9 0.001 0.006 3.338 3.339 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.708 3.102 prepare_preconditioner 11 7.9 0.000 0.000 2.492 2.499 make_preconditioner 11 8.9 0.000 0.002 2.492 2.499 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.104 2.426 calculate_first_density_matrix 1 7.0 0.001 0.005 2.272 2.273 make_m2s 4572 13.5 0.054 0.056 2.110 2.241 mp_waitall_1 137007 16.6 1.716 2.236 1.716 2.236 rs_pw_transfer 974 11.9 0.009 0.009 1.988 2.179 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.022 2.153 apply_single 119 13.6 0.000 0.000 2.021 2.153 make_images 4572 14.5 0.270 0.332 2.002 2.132 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.110 2.119 density_rs2pw 119 9.7 0.004 0.004 1.925 2.106 calculate_dm_sparse 119 9.5 0.000 0.001 2.019 2.083 pw_transfer 1439 11.6 0.065 0.070 2.002 2.035 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.910 1.948 grid_integrate_task_list 119 12.3 1.807 1.906 1.807 1.906 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.877 1.900 jit_kernel_multiply 10 15.8 0.934 1.818 0.934 1.818 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.800 1.801 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.799 1.799 ot_diis_step 108 11.5 0.012 0.012 1.749 1.749 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.638 1.651 fft_wrap_pw1pw2_140 487 13.2 0.288 0.300 1.576 1.614 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.558 1.567 cp_fm_diag_elpa_base 50 14.0 1.524 1.542 1.556 1.565 potential_pw2rs 119 12.3 0.008 0.009 1.552 1.555 acc_transpose_blocks 27432 15.5 0.112 0.117 1.466 1.494 fft3d_ps 1201 14.6 0.559 0.605 1.348 1.374 grid_collocate_task_list 119 9.7 1.219 1.319 1.219 1.319 wfi_extrapolate 11 7.9 0.001 0.001 1.302 1.302 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.734 1.282 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.206 1.217 mp_sum_l 7207 12.9 0.919 1.163 0.919 1.163 cp_fm_upper_to_full 72 14.2 0.815 1.156 0.815 1.156 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.100 1.103 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.084 1.102 multiply_cannon_sync_h2d 27432 15.5 0.991 1.080 0.991 1.080 dbcsr_complete_redistribute 329 12.2 0.122 0.152 0.756 1.029 make_images_data 4572 15.5 0.045 0.049 0.824 0.960 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.699 0.904 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.874 mp_alltoall_d11v 2130 13.8 0.717 0.874 0.717 0.874 acc_transpose_blocks_kernels 27432 16.5 0.267 0.276 0.841 0.863 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.830 0.835 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.555 0.823 cp_fm_cholesky_invert 11 10.9 0.718 0.721 0.718 0.721 mp_alltoall_i22 627 13.8 0.422 0.711 0.422 0.711 mp_alltoall_z22v 1201 16.6 0.673 0.703 0.673 0.703 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.663000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=522.636364, yerr=3.282813 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 600.944640E+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 0.012 0.028 27.909 27.910 qs_mol_dyn_low 1 2.0 0.003 0.003 27.733 27.740 qs_forces 11 3.9 0.002 0.003 27.668 27.669 qs_energies 11 4.9 0.001 0.001 25.954 25.957 scf_env_do_scf 11 5.9 0.000 0.001 21.299 21.299 scf_env_do_scf_inner_loop 108 6.5 0.004 0.027 18.792 18.793 velocity_verlet 10 3.0 0.001 0.001 14.299 14.302 dbcsr_multiply_generic 2286 12.5 0.089 0.093 11.974 12.078 qs_scf_new_mos 108 7.5 0.001 0.001 10.804 10.832 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.804 10.831 ot_scf_mini 108 9.5 0.002 0.002 10.139 10.169 multiply_cannon 2286 13.5 0.229 0.236 9.520 9.873 multiply_cannon_loop 2286 14.5 0.329 0.340 8.627 8.839 multiply_cannon_multrec 9144 15.5 1.590 1.849 5.742 6.058 rebuild_ks_matrix 119 8.3 0.000 0.000 6.002 6.022 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.002 6.022 ot_mini 108 10.5 0.001 0.001 5.645 5.677 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.356 5.373 qs_ot_get_derivative 108 11.5 0.001 0.001 4.343 4.371 dbcsr_mm_accdrv_process 12550 15.8 3.137 3.738 4.050 4.124 sum_up_and_integrate 119 10.3 0.038 0.042 3.657 3.660 integrate_v_rspace 119 11.3 0.003 0.003 3.619 3.623 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.516 3.520 calculate_rho_elec 119 8.7 0.060 0.061 3.516 3.519 init_scf_run 11 5.9 0.000 0.001 3.216 3.216 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.216 3.216 qs_ot_get_p 119 10.4 0.001 0.001 2.716 2.765 init_scf_loop 11 6.9 0.000 0.000 2.488 2.489 calculate_first_density_matrix 1 7.0 0.000 0.000 2.084 2.084 pw_transfer 1439 11.6 0.066 0.069 2.040 2.049 mp_waitall_1 115863 16.7 1.450 1.991 1.450 1.991 density_rs2pw 119 9.7 0.004 0.004 1.829 1.968 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.947 1.957 grid_integrate_task_list 119 12.3 1.853 1.940 1.853 1.940 make_m2s 4572 13.5 0.034 0.036 1.705 1.861 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.805 1.807 rs_pw_transfer 974 11.9 0.008 0.008 1.662 1.785 make_images 4572 14.5 0.267 0.299 1.616 1.770 jit_kernel_multiply 10 15.7 0.875 1.730 0.875 1.730 prepare_preconditioner 11 7.9 0.000 0.000 1.705 1.709 make_preconditioner 11 8.9 0.000 0.000 1.705 1.709 calculate_dm_sparse 119 9.5 0.000 0.000 1.687 1.704 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.598 1.622 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.614 1.614 fft_wrap_pw1pw2_140 487 13.2 0.365 0.373 1.603 1.613 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.511 1.529 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.524 1.525 grid_collocate_task_list 119 9.7 1.273 1.394 1.273 1.394 potential_pw2rs 119 12.3 0.010 0.011 1.388 1.390 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.368 1.374 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.323 1.331 cp_fm_diag_elpa_base 50 14.0 1.296 1.313 1.321 1.330 ot_diis_step 108 11.5 0.012 0.013 1.291 1.291 fft3d_ps 1201 14.6 0.562 0.573 1.280 1.288 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.218 1.219 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.168 1.189 apply_single 119 13.6 0.000 0.000 1.168 1.189 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.110 1.125 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.116 1.122 wfi_extrapolate 11 7.9 0.001 0.001 1.085 1.085 hybrid_alltoall_any 4725 16.4 0.063 0.176 0.789 1.019 make_images_data 4572 15.5 0.039 0.042 0.805 0.985 acc_transpose_blocks 9144 15.5 0.037 0.038 0.916 0.927 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.924 cp_fm_cholesky_invert 11 10.9 0.863 0.866 0.863 0.866 mp_alltoall_d11v 2130 13.8 0.770 0.864 0.770 0.864 multiply_cannon_sync_h2d 9144 15.5 0.717 0.795 0.717 0.795 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.361 0.778 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.768 0.771 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.671 0.723 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.712 0.721 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.677 0.685 mp_allgather_i34 2286 14.5 0.227 0.643 0.227 0.643 mp_alltoall_z22v 1201 16.6 0.592 0.626 0.592 0.626 jit_kernel_transpose 5 15.6 0.560 0.570 0.560 0.570 qs_create_task_list 11 7.9 0.001 0.001 0.542 0.567 generate_qs_task_list 11 8.9 0.187 0.210 0.541 0.566 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.910000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=570.909091, yerr=3.203820 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 768.659456E+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 0.017 0.034 41.524 41.525 qs_mol_dyn_low 1 2.0 0.003 0.003 41.251 41.258 qs_forces 11 3.9 0.003 0.003 40.538 40.542 qs_energies 11 4.9 0.001 0.001 38.556 38.562 scf_env_do_scf 11 5.9 0.001 0.001 32.884 32.884 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.139 25.140 velocity_verlet 10 3.0 0.001 0.001 23.102 23.115 dbcsr_multiply_generic 2286 12.5 0.096 0.099 17.171 17.332 qs_scf_new_mos 108 7.5 0.001 0.001 15.471 15.567 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.470 15.566 ot_scf_mini 108 9.5 0.002 0.002 14.407 14.510 multiply_cannon 2286 13.5 0.303 0.321 13.480 14.375 multiply_cannon_loop 2286 14.5 0.342 0.347 12.227 13.151 ot_mini 108 10.5 0.001 0.001 8.598 8.714 multiply_cannon_multrec 9144 15.5 3.398 4.770 8.507 8.590 init_scf_loop 11 6.9 0.000 0.000 7.719 7.721 rebuild_ks_matrix 119 8.3 0.000 0.000 7.237 7.378 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.237 7.378 prepare_preconditioner 11 7.9 0.000 0.000 6.755 6.768 make_preconditioner 11 8.9 0.000 0.000 6.755 6.768 qs_ot_get_derivative 108 11.5 0.001 0.001 6.581 6.682 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.541 6.669 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.300 6.650 dbcsr_mm_accdrv_process 12550 15.8 4.136 5.597 4.986 6.316 cp_fm_upper_to_full 72 14.2 3.154 4.557 3.154 4.557 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.194 4.200 calculate_rho_elec 119 8.7 0.118 0.121 4.193 4.199 sum_up_and_integrate 119 10.3 0.064 0.066 3.959 3.963 integrate_v_rspace 119 11.3 0.003 0.004 3.895 3.900 init_scf_run 11 5.9 0.000 0.001 3.632 3.632 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.632 3.632 qs_ot_get_p 119 10.4 0.001 0.001 3.101 3.233 mp_waitall_1 94719 16.7 2.228 3.205 2.228 3.205 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.400 2.826 dbcsr_complete_redistribute 329 12.2 0.290 0.297 1.967 2.780 pw_transfer 1439 11.6 0.069 0.069 2.635 2.641 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.536 2.542 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.672 2.485 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.195 2.448 apply_single 119 13.6 0.000 0.000 2.195 2.448 make_m2s 4572 13.5 0.038 0.038 2.173 2.351 mp_alltoall_i22 627 13.8 1.489 2.313 1.489 2.313 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.450 2.256 make_images 4572 14.5 0.351 0.383 2.052 2.229 calculate_first_density_matrix 1 7.0 0.000 0.000 2.218 2.220 density_rs2pw 119 9.7 0.004 0.004 2.135 2.162 fft_wrap_pw1pw2_140 487 13.2 0.617 0.619 2.148 2.155 calculate_dm_sparse 119 9.5 0.000 0.000 2.132 2.151 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.253 2.129 grid_integrate_task_list 119 12.3 2.047 2.055 2.047 2.055 ot_diis_step 108 11.5 0.014 0.015 1.994 1.994 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.922 1.922 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.902 1.903 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.776 1.780 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.630 1.681 mp_sum_l 7207 12.9 0.980 1.667 0.980 1.667 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.618 1.647 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.628 1.629 fft3d_ps 1201 14.6 0.597 0.610 1.553 1.560 grid_collocate_task_list 119 9.7 1.481 1.509 1.481 1.509 rs_pw_transfer 974 11.9 0.009 0.009 1.434 1.459 cp_fm_cholesky_invert 11 10.9 1.406 1.410 1.406 1.410 potential_pw2rs 119 12.3 0.014 0.014 1.368 1.373 wfi_extrapolate 11 7.9 0.001 0.001 1.352 1.352 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.333 1.333 cp_fm_diag_elpa_base 50 14.0 1.189 1.241 1.331 1.331 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.067 1.327 make_images_data 4572 15.5 0.043 0.046 1.022 1.243 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.213 1.221 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.118 1.134 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.114 1.132 multiply_cannon_sync_h2d 9144 15.5 1.043 1.046 1.043 1.046 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.985 1.037 mp_alltoall_d11v 2130 13.8 1.017 1.031 1.017 1.031 jit_kernel_multiply 6 15.8 0.822 1.011 0.822 1.011 qs_create_task_list 11 7.9 0.013 0.026 0.958 0.969 generate_qs_task_list 11 8.9 0.368 0.386 0.945 0.968 acc_transpose_blocks 9144 15.5 0.038 0.038 0.930 0.936 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.908 0.921 mp_alltoall_z22v 1201 16.6 0.820 0.839 0.820 0.839 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.525000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=718.090909, yerr=15.523909 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 500.776960E+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 1882585. 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 0.021 0.043 83.137 83.138 qs_mol_dyn_low 1 2.0 0.003 0.004 82.683 82.692 qs_forces 11 3.9 0.004 0.006 81.483 81.484 qs_energies 11 4.9 0.005 0.036 78.577 78.590 scf_env_do_scf 11 5.9 0.000 0.001 69.187 69.189 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 63.518 63.518 dbcsr_multiply_generic 2055 12.4 0.105 0.108 49.883 50.119 qs_scf_new_mos 99 7.5 0.000 0.001 45.904 46.033 qs_scf_loop_do_ot 99 8.5 0.001 0.001 45.903 46.033 velocity_verlet 10 3.0 0.001 0.001 43.702 43.715 ot_scf_mini 99 9.5 0.002 0.002 43.622 43.684 multiply_cannon 2055 13.4 0.186 0.192 41.789 42.510 multiply_cannon_loop 2055 14.4 1.498 1.530 40.767 41.531 ot_mini 99 10.5 0.001 0.001 25.562 25.634 qs_ot_get_derivative 99 11.5 0.001 0.001 18.832 18.914 multiply_cannon_multrec 49320 15.4 12.581 13.311 17.399 18.232 rebuild_ks_matrix 110 8.3 0.000 0.000 14.756 14.860 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.756 14.860 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.937 13.023 multiply_cannon_sync_h2d 49320 15.4 10.435 11.110 10.435 11.110 mp_waitall_1 220248 16.4 10.242 11.041 10.242 11.041 qs_ot_get_p 110 10.4 0.001 0.001 9.675 9.738 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.200 7.702 apply_single 110 13.6 0.000 0.001 7.200 7.702 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.107 7.665 sum_up_and_integrate 110 10.3 0.036 0.042 7.255 7.270 integrate_v_rspace 110 11.3 0.003 0.004 7.218 7.242 init_scf_run 11 5.9 0.000 0.002 7.156 7.157 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.156 7.156 multiply_cannon_metrocomm3 49320 15.4 0.080 0.083 6.067 7.089 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.740 6.880 calculate_rho_elec 110 8.6 0.021 0.026 6.739 6.880 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.612 6.644 ot_diis_step 99 11.5 0.006 0.006 6.560 6.560 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.788 5.788 init_scf_loop 11 6.9 0.000 0.001 5.642 5.642 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.308 5.369 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.202 5.237 cp_fm_diag_elpa_base 48 14.0 5.187 5.223 5.199 5.235 dbcsr_mm_accdrv_process 87628 16.1 1.884 1.955 4.691 4.955 rs_pw_transfer 902 11.9 0.012 0.013 3.922 4.632 mp_sum_l 6514 12.8 3.689 4.500 3.689 4.500 density_rs2pw 110 9.6 0.004 0.005 3.549 4.155 wfi_extrapolate 11 7.9 0.001 0.001 4.136 4.136 make_m2s 4110 13.4 0.060 0.064 3.960 4.078 make_images 4110 14.4 0.177 0.190 3.865 3.986 calculate_dm_sparse 110 9.5 0.001 0.001 3.767 3.890 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.696 3.700 prepare_preconditioner 11 7.9 0.000 0.000 3.450 3.474 make_preconditioner 11 8.9 0.000 0.000 3.450 3.474 grid_integrate_task_list 110 12.3 3.247 3.421 3.247 3.421 pw_transfer 1331 11.6 0.055 0.070 3.255 3.331 multiply_cannon_metrocomm1 49320 15.4 0.061 0.064 2.196 3.326 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.269 3.304 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.239 3.283 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.166 3.244 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.100 3.136 calculate_first_density_matrix 1 7.0 0.000 0.001 2.928 2.935 potential_pw2rs 110 12.3 0.006 0.007 2.761 2.785 fft_wrap_pw1pw2_140 451 13.1 0.453 0.497 2.684 2.770 jit_kernel_multiply 13 15.9 2.536 2.651 2.536 2.651 mp_alltoall_d11v 2046 13.8 2.081 2.597 2.081 2.597 mp_waitany 14300 13.8 1.852 2.596 1.852 2.596 fft3d_ps 1111 14.6 0.795 0.878 2.360 2.407 grid_collocate_task_list 110 9.6 2.082 2.325 2.082 2.325 acc_transpose_blocks 49320 15.4 0.208 0.218 2.068 2.141 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.950 1.967 mp_sum_d 3889 11.9 1.430 1.943 1.430 1.943 make_images_data 4110 15.4 0.042 0.045 1.801 1.937 cp_fm_cholesky_invert 11 10.9 1.912 1.916 1.912 1.916 hybrid_alltoall_any 4261 16.3 0.082 0.481 1.558 1.809 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.761 1.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.138000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.818182, yerr=2.081004 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 586.911744E+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 1934587. 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 0.053 0.088 69.508 69.509 qs_mol_dyn_low 1 2.0 0.003 0.004 69.126 69.135 qs_forces 11 3.9 0.004 0.007 68.717 68.718 qs_energies 11 4.9 0.001 0.002 65.379 65.383 scf_env_do_scf 11 5.9 0.000 0.001 56.832 56.835 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.182 49.183 dbcsr_multiply_generic 2055 12.4 0.113 0.118 37.445 37.601 velocity_verlet 10 3.0 0.001 0.001 36.638 36.646 qs_scf_new_mos 99 7.5 0.001 0.001 32.726 32.871 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.726 32.871 multiply_cannon 2055 13.4 0.221 0.240 30.949 32.062 ot_scf_mini 99 9.5 0.003 0.003 31.075 31.226 multiply_cannon_loop 2055 14.4 0.921 0.943 29.675 30.388 ot_mini 99 10.5 0.001 0.001 18.035 18.193 multiply_cannon_multrec 24660 15.4 7.663 9.162 13.838 15.333 rebuild_ks_matrix 110 8.3 0.000 0.000 13.875 13.947 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.874 13.946 qs_ot_get_derivative 99 11.5 0.001 0.010 12.256 12.414 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.220 12.283 mp_waitall_1 176588 16.5 7.583 9.774 7.583 9.774 multiply_cannon_sync_h2d 24660 15.4 7.056 8.006 7.056 8.006 multiply_cannon_metrocomm3 24660 15.4 0.067 0.069 5.090 7.628 init_scf_loop 11 6.9 0.000 0.000 7.615 7.615 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.505 7.124 apply_single 110 13.6 0.000 0.001 6.504 7.124 sum_up_and_integrate 110 10.3 0.052 0.059 6.726 6.739 integrate_v_rspace 110 11.3 0.003 0.003 6.674 6.689 qs_ot_get_p 110 10.4 0.001 0.001 6.413 6.607 dbcsr_mm_accdrv_process 52282 16.1 4.676 5.650 6.014 6.295 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.232 6.239 calculate_rho_elec 110 8.6 0.040 0.048 6.231 6.239 init_scf_run 11 5.9 0.000 0.001 6.156 6.156 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.156 6.156 ot_diis_step 99 11.5 0.010 0.012 5.731 5.731 prepare_preconditioner 11 7.9 0.000 0.000 5.584 5.605 make_preconditioner 11 8.9 0.000 0.000 5.584 5.605 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.631 5.365 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.180 5.330 make_m2s 4110 13.4 0.056 0.060 4.147 4.623 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.579 4.600 make_images 4110 14.4 0.397 0.439 4.040 4.512 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.155 4.155 density_rs2pw 110 9.6 0.004 0.005 3.351 3.871 pw_transfer 1331 11.6 0.067 0.074 3.676 3.811 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.568 3.707 rs_pw_transfer 902 11.9 0.012 0.014 3.101 3.651 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.586 3.603 cp_fm_diag_elpa_base 48 14.0 3.540 3.561 3.583 3.600 wfi_extrapolate 11 7.9 0.001 0.001 3.535 3.535 grid_integrate_task_list 110 12.3 3.151 3.363 3.151 3.363 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.309 3.311 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.135 3.213 fft_wrap_pw1pw2_140 451 13.1 0.519 0.534 3.035 3.173 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.960 3.017 calculate_dm_sparse 110 9.5 0.001 0.001 2.944 2.974 make_images_data 4110 15.4 0.046 0.050 2.338 2.808 hybrid_alltoall_any 4261 16.3 0.102 0.444 2.069 2.791 fft3d_ps 1111 14.6 1.108 1.327 2.545 2.693 cp_fm_cholesky_invert 11 10.9 2.576 2.583 2.576 2.583 calculate_first_density_matrix 1 7.0 0.000 0.000 2.532 2.535 potential_pw2rs 110 12.3 0.008 0.009 2.502 2.518 grid_collocate_task_list 110 9.6 2.056 2.484 2.056 2.484 mp_sum_l 6514 12.8 1.730 2.409 1.730 2.409 jit_kernel_multiply 12 16.3 0.988 2.294 0.988 2.294 mp_alltoall_d11v 2046 13.8 1.778 2.002 1.778 2.002 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.936 1.952 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.880 1.882 mp_waitany 10164 13.8 1.269 1.832 1.269 1.832 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.810 1.822 multiply_cannon_metrocomm4 22605 15.4 0.075 0.080 0.781 1.647 mp_allgather_i34 2055 14.4 0.604 1.625 0.604 1.625 acc_transpose_blocks 24660 15.4 0.109 0.112 1.560 1.586 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.216 0.996 1.538 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.526 1.535 mp_irecv_dv 57340 16.2 0.655 1.528 0.655 1.528 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.371 1.478 dbcsr_complete_redistribute 325 12.2 0.245 0.308 1.139 1.404 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.509000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.818182, yerr=5.982067 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 659.525632E+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 0.020 0.053 60.130 60.132 qs_mol_dyn_low 1 2.0 0.003 0.004 59.769 59.778 qs_forces 11 3.9 0.004 0.008 59.672 59.674 qs_energies 11 4.9 0.002 0.007 56.457 56.459 scf_env_do_scf 11 5.9 0.001 0.001 48.434 48.435 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 39.881 39.882 velocity_verlet 10 3.0 0.001 0.001 32.667 32.670 dbcsr_multiply_generic 2055 12.4 0.105 0.110 28.256 28.473 qs_scf_new_mos 99 7.5 0.001 0.001 24.733 24.823 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.732 24.823 ot_scf_mini 99 9.5 0.002 0.003 23.519 23.628 multiply_cannon 2055 13.4 0.213 0.224 22.024 23.285 multiply_cannon_loop 2055 14.4 0.615 0.626 20.853 21.969 ot_mini 99 10.5 0.001 0.001 13.354 13.462 rebuild_ks_matrix 110 8.3 0.000 0.000 12.450 12.573 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.450 12.573 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.965 11.078 multiply_cannon_multrec 16440 15.4 3.973 5.036 9.718 10.765 mp_waitall_1 139946 16.5 6.858 9.868 6.858 9.868 qs_ot_get_derivative 99 11.5 0.001 0.001 8.969 9.074 init_scf_loop 11 6.9 0.001 0.006 8.521 8.522 multiply_cannon_metrocomm3 16440 15.4 0.042 0.044 4.172 7.043 prepare_preconditioner 11 7.9 0.000 0.001 6.738 6.753 make_preconditioner 11 8.9 0.000 0.002 6.738 6.752 sum_up_and_integrate 110 10.3 0.060 0.061 6.594 6.609 integrate_v_rspace 110 11.3 0.003 0.004 6.533 6.548 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.087 6.433 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.031 6.043 calculate_rho_elec 110 8.6 0.059 0.060 6.031 6.042 dbcsr_mm_accdrv_process 34862 16.1 4.667 5.251 5.599 5.739 init_scf_run 11 5.9 0.000 0.001 5.558 5.558 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.558 5.558 qs_ot_get_p 110 10.4 0.001 0.001 5.364 5.485 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.926 5.351 apply_single 110 13.6 0.000 0.000 4.926 5.351 make_m2s 4110 13.4 0.050 0.051 4.154 4.520 multiply_cannon_sync_h2d 16440 15.4 3.715 4.477 3.715 4.477 make_images 4110 14.4 0.390 0.509 4.039 4.405 density_rs2pw 110 9.6 0.005 0.006 3.148 4.389 ot_diis_step 99 11.5 0.011 0.011 4.359 4.359 rs_pw_transfer 902 11.9 0.010 0.012 2.767 4.008 qs_ot_p2m_diag 48 11.0 0.041 0.044 3.751 3.756 pw_transfer 1331 11.6 0.066 0.072 3.702 3.713 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.083 3.711 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.595 3.609 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.414 3.414 grid_integrate_task_list 110 12.3 3.166 3.411 3.166 3.411 fft_wrap_pw1pw2_140 451 13.1 0.636 0.647 3.081 3.093 wfi_extrapolate 11 7.9 0.001 0.001 2.964 2.964 make_images_data 4110 15.4 0.043 0.048 2.430 2.916 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.878 2.887 cp_fm_diag_elpa_base 48 14.0 2.811 2.842 2.875 2.884 hybrid_alltoall_any 4261 16.3 0.105 0.375 2.165 2.849 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.810 2.812 cp_fm_cholesky_invert 11 10.9 2.652 2.657 2.652 2.657 calculate_dm_sparse 110 9.5 0.001 0.001 2.519 2.550 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.489 2.536 mp_waitany 17072 13.8 1.222 2.530 1.222 2.530 calculate_first_density_matrix 1 7.0 0.001 0.005 2.503 2.504 grid_collocate_task_list 110 9.6 2.083 2.462 2.083 2.462 fft3d_ps 1111 14.6 1.093 1.104 2.419 2.429 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.295 2.348 potential_pw2rs 110 12.3 0.011 0.011 2.325 2.335 multiply_cannon_metrocomm4 14385 15.4 0.046 0.050 0.871 2.329 mp_irecv_dv 48980 15.7 0.801 2.204 0.801 2.204 rs_pw_transfer_RS2PW_140 121 11.5 0.183 0.196 0.942 2.181 mp_alltoall_d11v 2046 13.8 1.711 2.164 1.711 2.164 mp_sum_l 6514 12.8 1.471 2.075 1.471 2.075 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.030 2.032 jit_kernel_multiply 10 16.7 0.545 1.923 0.545 1.923 dbcsr_complete_redistribute 325 12.2 0.323 0.348 1.423 1.891 cp_fm_upper_to_full 70 14.2 1.385 1.757 1.385 1.757 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.679 1.693 cp_fm_cholesky_decompose 22 10.9 1.593 1.611 1.593 1.611 mp_allgather_i34 2055 14.4 0.494 1.606 0.494 1.606 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.486 1.500 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.366 1.471 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.945 1.393 rs_gather_matrices 110 12.3 0.234 0.260 0.963 1.300 acc_transpose_blocks 16440 15.4 0.072 0.073 1.254 1.275 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.232 1.237 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.123 1.212 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.132000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.363636, yerr=9.048081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 725.700608E+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 0.037 0.152 65.319 65.320 qs_mol_dyn_low 1 2.0 0.003 0.006 64.907 64.919 qs_forces 11 3.9 0.003 0.005 64.796 64.797 qs_energies 11 4.9 0.002 0.007 61.372 61.377 scf_env_do_scf 11 5.9 0.001 0.001 53.127 53.130 scf_env_do_scf_inner_loop 99 6.5 0.005 0.009 41.569 41.569 velocity_verlet 10 3.0 0.001 0.002 36.874 36.878 dbcsr_multiply_generic 2055 12.4 0.111 0.117 29.437 29.676 qs_scf_new_mos 99 7.5 0.001 0.001 26.499 26.601 qs_scf_loop_do_ot 99 8.5 0.001 0.002 26.498 26.600 ot_scf_mini 99 9.5 0.010 0.062 24.883 24.994 multiply_cannon 2055 13.4 0.244 0.270 22.618 23.912 multiply_cannon_loop 2055 14.4 0.884 0.908 21.093 21.602 ot_mini 99 10.5 0.004 0.024 14.169 14.309 multiply_cannon_multrec 24660 15.4 4.231 6.587 12.648 13.939 rebuild_ks_matrix 110 8.3 0.000 0.000 12.207 12.360 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.207 12.360 init_scf_loop 11 6.9 0.001 0.004 11.520 11.520 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.774 10.908 qs_ot_get_derivative 99 11.5 0.001 0.002 10.020 10.147 prepare_preconditioner 11 7.9 0.000 0.000 9.765 9.781 make_preconditioner 11 8.9 0.000 0.001 9.765 9.781 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.986 9.457 dbcsr_mm_accdrv_process 52304 16.0 6.802 8.478 8.269 9.180 sum_up_and_integrate 110 10.3 0.068 0.073 6.492 6.508 integrate_v_rspace 110 11.3 0.003 0.003 6.424 6.436 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.146 6.157 calculate_rho_elec 110 8.6 0.078 0.082 6.145 6.157 mp_waitall_1 121746 16.5 4.201 6.130 4.201 6.130 qs_ot_get_p 110 10.4 0.001 0.001 5.720 5.880 make_m2s 4110 13.4 0.060 0.062 5.183 5.677 make_images 4110 14.4 0.574 0.693 5.042 5.533 init_scf_run 11 5.9 0.000 0.001 5.469 5.470 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.469 5.470 cp_fm_upper_to_full 70 14.2 3.362 4.877 3.362 4.877 ot_diis_step 99 11.5 0.011 0.013 4.112 4.112 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.010 4.101 apply_single 110 13.6 0.000 0.000 4.010 4.101 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.975 3.990 dbcsr_complete_redistribute 325 12.2 0.420 0.460 2.662 3.797 pw_transfer 1331 11.6 0.066 0.075 3.744 3.775 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.638 3.672 density_rs2pw 110 9.6 0.004 0.005 3.011 3.519 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.507 3.508 multiply_cannon_metrocomm3 24660 15.4 0.036 0.036 1.472 3.461 grid_integrate_task_list 110 12.3 3.262 3.434 3.262 3.434 multiply_cannon_sync_h2d 24660 15.4 3.201 3.428 3.201 3.428 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.335 3.402 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.112 3.238 fft_wrap_pw1pw2_140 451 13.1 0.667 0.691 3.129 3.167 hybrid_alltoall_any 4261 16.3 0.120 0.454 2.273 3.108 make_images_data 4110 15.4 0.046 0.050 2.582 3.069 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.957 3.020 wfi_extrapolate 11 7.9 0.001 0.002 2.975 2.975 calculate_dm_sparse 110 9.5 0.001 0.001 2.898 2.930 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.916 2.928 cp_fm_diag_elpa_base 48 14.0 2.770 2.828 2.913 2.926 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.771 2.881 rs_pw_transfer 902 11.9 0.010 0.011 2.358 2.877 mp_alltoall_i22 605 13.7 1.663 2.842 1.663 2.842 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.785 2.787 cp_fm_cholesky_invert 11 10.9 2.594 2.602 2.594 2.602 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.428 2.473 grid_collocate_task_list 110 9.6 2.181 2.438 2.181 2.438 fft3d_ps 1111 14.6 1.087 1.120 2.402 2.422 calculate_first_density_matrix 1 7.0 0.001 0.006 2.399 2.402 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.281 2.282 potential_pw2rs 110 12.3 0.012 0.013 2.140 2.146 mp_alltoall_d11v 2046 13.8 1.837 2.026 1.837 2.026 jit_kernel_multiply 10 15.7 1.138 1.847 1.138 1.847 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.675 1.710 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.602 1.705 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.674 1.690 cp_fm_cholesky_decompose 22 10.9 1.616 1.664 1.616 1.664 mp_allgather_i34 2055 14.4 0.626 1.603 0.626 1.603 acc_transpose_blocks 24660 15.4 0.106 0.108 1.519 1.567 mp_waitany 13376 13.8 1.051 1.555 1.051 1.555 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.518 1.536 multiply_cannon_metrocomm4 20550 15.4 0.061 0.065 0.839 1.476 mp_sum_l 6514 12.8 1.040 1.445 1.040 1.445 mp_irecv_dv 62702 16.1 0.735 1.392 0.735 1.392 rs_pw_transfer_RS2PW_140 121 11.5 0.171 0.178 0.820 1.353 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.221 1.330 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.320000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=688.818182, yerr=7.420098 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 831.549440E+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 0.014 0.030 65.008 65.009 qs_mol_dyn_low 1 2.0 0.003 0.003 64.766 64.774 qs_forces 11 3.9 0.003 0.003 64.182 64.183 qs_energies 11 4.9 0.001 0.001 60.374 60.381 scf_env_do_scf 11 5.9 0.000 0.001 51.332 51.332 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 41.246 41.248 velocity_verlet 10 3.0 0.001 0.001 36.567 36.602 dbcsr_multiply_generic 2055 12.4 0.107 0.111 25.147 25.404 qs_scf_new_mos 99 7.5 0.001 0.001 24.496 24.559 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.496 24.558 ot_scf_mini 99 9.5 0.002 0.002 23.134 23.165 multiply_cannon 2055 13.4 0.248 0.261 18.286 19.964 multiply_cannon_loop 2055 14.4 0.320 0.335 16.752 17.329 rebuild_ks_matrix 110 8.3 0.000 0.000 12.872 12.903 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.871 12.902 ot_mini 99 10.5 0.001 0.001 12.169 12.190 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.432 11.456 init_scf_loop 11 6.9 0.000 0.000 10.012 10.014 mp_waitall_1 103326 16.6 7.751 9.516 7.751 9.516 multiply_cannon_multrec 8220 15.4 3.191 4.529 7.411 8.421 qs_ot_get_derivative 99 11.5 0.001 0.001 8.244 8.278 prepare_preconditioner 11 7.9 0.000 0.000 8.189 8.194 make_preconditioner 11 8.9 0.000 0.000 8.189 8.194 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.690 7.825 sum_up_and_integrate 110 10.3 0.079 0.081 7.324 7.336 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.290 7.307 calculate_rho_elec 110 8.6 0.114 0.115 7.290 7.306 integrate_v_rspace 110 11.3 0.003 0.003 7.245 7.255 qs_ot_get_p 110 10.4 0.001 0.001 6.758 6.793 init_scf_run 11 5.9 0.000 0.003 5.657 5.657 scf_env_initial_rho_setup 11 6.9 0.002 0.002 5.656 5.657 make_m2s 4110 13.4 0.039 0.040 5.178 5.506 make_images 4110 14.4 0.635 0.690 5.047 5.377 dbcsr_mm_accdrv_process 17442 15.9 2.720 3.491 4.091 5.001 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.965 4.975 pw_transfer 1331 11.6 0.067 0.072 4.599 4.630 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.561 4.562 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.490 4.523 density_rs2pw 110 9.6 0.004 0.005 3.918 4.288 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 2.930 4.216 cp_fm_cholesky_invert 11 10.9 4.203 4.208 4.203 4.208 ot_diis_step 99 11.5 0.012 0.012 3.873 3.873 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.836 3.843 cp_fm_diag_elpa_base 48 14.0 3.758 3.802 3.834 3.840 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.795 3.833 apply_single 110 13.6 0.000 0.000 3.795 3.833 fft_wrap_pw1pw2_140 451 13.1 0.830 0.842 3.725 3.760 make_images_data 4110 15.4 0.038 0.044 2.889 3.590 grid_integrate_task_list 110 12.3 3.339 3.544 3.339 3.544 rs_pw_transfer 902 11.9 0.010 0.010 3.185 3.541 hybrid_alltoall_any 4261 16.3 0.200 0.864 2.689 3.321 wfi_extrapolate 11 7.9 0.001 0.001 3.253 3.253 fft3d_ps 1111 14.6 1.148 1.183 3.039 3.077 multiply_cannon_sync_h2d 8220 15.4 2.901 3.047 2.901 3.047 potential_pw2rs 110 12.3 0.015 0.016 2.735 2.759 qs_energies_init_hamiltonians 11 5.9 0.003 0.020 2.737 2.738 calculate_dm_sparse 110 9.5 0.001 0.001 2.613 2.654 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.619 2.623 grid_collocate_task_list 110 9.6 2.284 2.546 2.284 2.546 mp_alltoall_d11v 2046 13.8 2.150 2.512 2.150 2.512 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.397 2.415 cp_fm_cholesky_decompose 22 10.9 2.381 2.399 2.381 2.399 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.328 2.348 calculate_first_density_matrix 1 7.0 0.000 0.000 2.297 2.299 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 1.709 2.265 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.195 2.201 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.772 1.991 mp_allgather_i34 2055 14.4 0.717 1.946 0.717 1.946 dbcsr_complete_redistribute 325 12.2 0.552 0.582 1.761 1.880 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.772 1.777 qs_env_update_s_mstruct 11 6.9 0.028 0.040 1.566 1.704 mp_alltoall_z22v 1111 16.6 1.654 1.693 1.654 1.693 mp_waitany 9240 13.8 1.259 1.649 1.259 1.649 make_images_sizes 4110 15.4 0.005 0.005 0.877 1.562 mp_alltoall_i44 4110 16.4 0.872 1.557 0.872 1.557 mp_sum_l 6514 12.8 0.984 1.521 0.984 1.521 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.491 1.504 jit_kernel_multiply 8 16.0 1.063 1.366 1.063 1.366 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.326 1.349 rs_gather_matrices 110 12.3 0.326 0.368 1.062 1.347 qs_create_task_list 11 7.9 0.000 0.001 1.231 1.333 generate_qs_task_list 11 8.9 0.374 0.440 1.230 1.333 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=65.009000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=782.272727, yerr=8.975633 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 1.321095E+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 0.046 0.089 99.791 99.792 qs_mol_dyn_low 1 2.0 0.003 0.003 99.327 99.336 qs_forces 11 3.9 0.003 0.003 99.258 99.258 qs_energies 11 4.9 0.001 0.001 94.852 94.856 scf_env_do_scf 11 5.9 0.001 0.001 83.971 83.971 velocity_verlet 10 3.0 0.001 0.001 63.292 63.298 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 51.718 51.720 init_scf_loop 11 6.9 0.000 0.000 32.167 32.170 dbcsr_multiply_generic 2055 12.4 0.120 0.124 31.313 31.613 qs_scf_new_mos 99 7.5 0.001 0.001 31.022 31.076 qs_scf_loop_do_ot 99 8.5 0.001 0.001 31.021 31.076 prepare_preconditioner 11 7.9 0.000 0.000 29.936 29.949 make_preconditioner 11 8.9 0.000 0.000 29.936 29.949 make_full_inverse_cholesky 11 9.9 0.000 0.000 23.901 29.343 ot_scf_mini 99 9.5 0.002 0.002 29.127 29.155 multiply_cannon 2055 13.4 0.340 0.366 22.735 24.497 multiply_cannon_loop 2055 14.4 0.342 0.345 20.591 21.908 cp_fm_upper_to_full 70 14.2 12.907 18.593 12.907 18.593 rebuild_ks_matrix 110 8.3 0.001 0.001 15.600 15.730 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 15.599 15.729 ot_mini 99 10.5 0.001 0.001 15.543 15.614 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.090 14.205 dbcsr_complete_redistribute 325 12.2 1.028 1.056 7.902 11.369 mp_waitall_1 84994 16.7 9.324 11.149 9.324 11.149 qs_ot_get_derivative 99 11.5 0.001 0.001 10.794 10.819 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.695 10.165 multiply_cannon_multrec 8220 15.4 4.417 4.611 9.621 9.712 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.019 9.456 mp_alltoall_i22 605 13.7 5.642 9.169 5.642 9.169 qs_rho_update_rho_low 110 7.6 0.001 0.001 9.115 9.155 calculate_rho_elec 110 8.6 0.225 0.226 9.114 9.155 sum_up_and_integrate 110 10.3 0.151 0.152 8.491 8.505 integrate_v_rspace 110 11.3 0.003 0.004 8.340 8.354 qs_ot_get_p 110 10.4 0.001 0.001 7.924 8.014 cp_fm_cholesky_invert 11 10.9 7.347 7.352 7.347 7.352 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 5.564 6.899 make_m2s 4110 13.4 0.043 0.044 6.187 6.653 make_images 4110 14.4 0.880 0.932 5.996 6.459 init_scf_run 11 5.9 0.000 0.001 6.452 6.453 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.452 6.452 pw_transfer 1331 11.6 0.075 0.075 6.235 6.258 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 6.119 6.142 qs_ot_p2m_diag 48 11.0 0.151 0.155 5.811 5.819 dbcsr_mm_accdrv_process 11614 15.7 3.322 3.666 5.060 5.301 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.285 5.286 fft_wrap_pw1pw2_140 451 13.1 1.336 1.340 5.146 5.172 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.745 5.119 apply_single 110 13.6 0.000 0.000 4.745 5.119 density_rs2pw 110 9.6 0.004 0.004 4.970 5.021 ot_diis_step 99 11.5 0.015 0.016 4.669 4.669 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.486 4.486 cp_fm_diag_elpa_base 48 14.0 3.911 4.115 4.484 4.484 hybrid_alltoall_any 4261 16.3 0.256 0.554 3.275 4.144 make_images_data 4110 15.4 0.042 0.045 3.331 4.138 fft3d_ps 1111 14.6 1.311 1.327 4.050 4.075 multiply_cannon_sync_h2d 8220 15.4 3.952 3.958 3.952 3.958 wfi_extrapolate 11 7.9 0.001 0.001 3.907 3.907 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.442 3.858 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.752 3.753 grid_integrate_task_list 110 12.3 3.676 3.736 3.676 3.736 rs_pw_transfer 902 11.9 0.010 0.011 3.510 3.561 potential_pw2rs 110 12.3 0.021 0.021 3.367 3.380 calculate_dm_sparse 110 9.5 0.001 0.001 3.308 3.368 cp_fm_cholesky_decompose 22 10.9 3.097 3.124 3.097 3.124 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.048 3.051 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.978 2.991 mp_alltoall_d11v 2046 13.8 2.666 2.796 2.666 2.796 grid_collocate_task_list 110 9.6 2.645 2.679 2.645 2.679 mp_alltoall_z22v 1111 16.6 2.485 2.505 2.485 2.505 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.430 2.470 mp_sum_l 6514 12.8 1.512 2.452 1.512 2.452 calculate_first_density_matrix 1 7.0 0.000 0.000 2.398 2.399 qs_env_update_s_mstruct 11 6.9 0.021 0.031 2.268 2.338 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.093 2.145 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.113 2.127 copy_dbcsr_to_fm 151 11.3 0.003 0.003 2.015 2.053 mp_allgather_i34 2055 14.4 0.912 2.019 0.912 2.019 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=99.792000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1209.545455, yerr=40.902020 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 633.057280E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57511. MP_Allreduce 11049 800. MP_Sync 87 MP_Alltoall 2226 2310002. 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 0.025 0.080 224.517 224.532 qs_mol_dyn_low 1 2.0 0.013 0.045 223.937 223.950 qs_forces 11 3.9 0.011 0.035 223.807 223.811 qs_energies 11 4.9 0.001 0.003 217.976 217.995 scf_env_do_scf 11 5.9 0.001 0.001 200.382 200.385 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 176.447 176.449 qs_scf_new_mos 117 7.6 0.001 0.001 132.993 133.188 qs_scf_loop_do_ot 117 8.6 0.001 0.001 132.993 133.188 velocity_verlet 10 3.0 0.001 0.001 132.010 132.011 dbcsr_multiply_generic 2507 12.6 0.184 0.219 127.083 127.738 ot_scf_mini 117 9.6 0.003 0.003 126.317 126.533 multiply_cannon 2507 13.6 0.242 0.251 102.219 104.202 multiply_cannon_loop 2507 14.6 2.072 2.116 99.717 101.750 ot_mini 117 10.6 0.001 0.001 67.964 68.179 multiply_cannon_multrec 60168 15.6 33.433 35.265 41.669 43.778 qs_ot_get_derivative 117 11.6 0.001 0.001 42.936 43.132 qs_ot_get_p 128 10.4 0.001 0.001 36.169 36.480 rebuild_ks_matrix 128 8.3 0.001 0.001 35.891 36.301 qs_ks_build_kohn_sham_matrix 128 9.3 0.030 0.136 35.890 36.300 mp_waitall_1 267128 16.5 31.053 34.122 31.053 34.122 qs_ks_update_qs_env 128 7.6 0.001 0.001 32.286 32.628 qs_ot_p2m_diag 83 11.4 0.079 0.091 28.887 28.935 multiply_cannon_sync_h2d 60168 15.6 27.326 28.914 27.326 28.914 cp_dbcsr_syevd 83 12.4 0.004 0.005 26.209 26.210 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.533 25.146 apply_single 128 13.6 0.001 0.001 24.533 25.146 ot_diis_step 117 11.6 0.008 0.008 24.744 24.746 init_scf_loop 11 6.9 0.000 0.000 23.855 23.857 cp_fm_diag_elpa 83 13.4 0.000 0.000 22.384 22.421 cp_fm_diag_elpa_base 83 14.4 22.266 22.314 22.379 22.417 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.607 20.772 prepare_preconditioner 11 7.9 0.000 0.000 18.827 18.894 make_preconditioner 11 8.9 0.000 0.000 18.827 18.894 multiply_cannon_metrocomm3 60168 15.6 0.114 0.119 16.082 18.346 make_full_inverse_cholesky 11 9.9 0.000 0.000 18.029 18.178 sum_up_and_integrate 128 10.3 0.089 0.108 15.940 15.956 integrate_v_rspace 128 11.3 0.005 0.008 15.850 15.869 qs_rho_update_rho_low 128 7.7 0.001 0.004 15.604 15.698 calculate_rho_elec 128 8.7 0.046 0.065 15.603 15.697 make_m2s 5014 13.6 0.104 0.112 15.058 15.436 make_images 5014 14.6 0.402 0.421 14.876 15.266 init_scf_run 11 5.9 0.000 0.001 13.213 13.214 scf_env_initial_rho_setup 11 6.9 0.001 0.023 13.213 13.214 density_rs2pw 128 9.7 0.006 0.007 8.974 12.242 rs_pw_transfer 1046 11.9 0.016 0.019 8.395 11.753 cp_fm_cholesky_invert 11 10.9 11.160 11.168 11.160 11.168 mp_sum_l 7870 13.0 8.591 9.574 8.591 9.574 wfi_extrapolate 11 7.9 0.001 0.001 9.572 9.572 pw_transfer 1547 11.6 0.075 0.092 8.789 9.065 multiply_cannon_metrocomm1 60168 15.6 0.089 0.093 6.673 8.881 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 8.585 8.867 calculate_dm_sparse 128 9.5 0.001 0.001 8.584 8.687 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.281 8.419 make_images_data 5014 15.6 0.065 0.071 7.387 8.370 dbcsr_mm_accdrv_process 124484 16.2 3.158 3.263 7.796 8.289 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.955 8.023 hybrid_alltoall_any 5200 16.5 0.292 2.260 6.396 7.641 grid_integrate_task_list 128 12.3 6.965 7.575 6.965 7.575 fft_wrap_pw1pw2_140 523 13.2 1.256 1.317 6.862 7.141 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.005 6.885 6.894 fft3d_ps 1291 14.7 2.151 2.791 6.370 6.803 mp_alltoall_d11v 2415 14.1 4.634 6.795 4.634 6.795 potential_pw2rs 128 12.3 0.009 0.011 6.257 6.365 mp_waitany 16020 13.9 2.920 6.180 2.920 6.180 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.947 6.092 grid_collocate_task_list 128 9.7 4.544 5.766 4.544 5.766 rs_pw_transfer_RS2PW_140 139 11.5 0.282 0.296 2.240 5.542 cp_fm_cholesky_decompose 22 10.9 5.373 5.387 5.373 5.387 rs_gather_matrices 128 12.3 0.366 0.409 2.467 4.586 mp_sum_d 4479 12.1 3.779 4.530 3.779 4.530 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=224.532000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.181818, yerr=6.779892 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/16/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 2.199914E+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 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 834.785280E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57909. MP_Allreduce 11058 964. MP_Sync 87 MP_Alltoall 1983 4959973. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.031 209.560 209.561 qs_mol_dyn_low 1 2.0 0.003 0.003 209.067 209.079 qs_forces 11 3.9 0.010 0.014 208.980 208.981 qs_energies 11 4.9 0.001 0.001 202.027 202.055 scf_env_do_scf 11 5.9 0.001 0.001 184.982 184.992 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 147.729 147.731 velocity_verlet 10 3.0 0.001 0.001 130.748 130.749 qs_scf_new_mos 118 7.6 0.001 0.001 104.357 104.849 qs_scf_loop_do_ot 118 8.6 0.001 0.001 104.356 104.848 dbcsr_multiply_generic 2527 12.6 0.191 0.198 102.615 104.648 ot_scf_mini 118 9.6 0.004 0.004 99.348 99.818 multiply_cannon 2527 13.6 0.484 0.534 79.562 85.380 multiply_cannon_loop 2527 14.6 1.259 1.299 75.624 80.720 ot_mini 118 10.6 0.001 0.001 53.594 54.023 mp_waitall_1 216438 16.6 27.942 46.021 27.942 46.021 init_scf_loop 11 6.9 0.000 0.000 37.158 37.159 multiply_cannon_multrec 30324 15.6 21.856 27.382 31.480 37.146 rebuild_ks_matrix 129 8.3 0.001 0.001 35.359 35.811 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 35.358 35.811 multiply_cannon_metrocomm3 30324 15.6 0.092 0.100 17.304 35.423 prepare_preconditioner 11 7.9 0.000 0.000 32.548 32.595 make_preconditioner 11 8.9 0.000 0.000 32.548 32.595 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.895 32.322 make_full_inverse_cholesky 11 9.9 0.000 0.000 31.125 31.765 qs_ot_get_derivative 118 11.6 0.001 0.002 31.118 31.576 qs_ot_get_p 129 10.4 0.001 0.001 26.700 27.294 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.347 23.933 apply_single 129 13.6 0.001 0.001 22.347 23.933 ot_diis_step 118 11.6 0.015 0.015 22.199 22.201 multiply_cannon_sync_h2d 30324 15.6 19.126 21.900 19.126 21.900 qs_ot_p2m_diag 83 11.4 0.187 0.216 21.246 21.358 cp_dbcsr_syevd 83 12.4 0.005 0.006 19.769 19.771 cp_fm_cholesky_invert 11 10.9 19.091 19.108 19.091 19.108 make_m2s 5054 13.6 0.090 0.095 15.633 17.286 make_images 5054 14.6 1.159 1.343 15.423 17.074 sum_up_and_integrate 129 10.3 0.117 0.134 16.232 16.261 integrate_v_rspace 129 11.3 0.004 0.004 16.114 16.149 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.046 16.072 cp_fm_diag_elpa_base 83 14.4 15.750 15.870 16.039 16.068 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.655 15.734 calculate_rho_elec 129 8.7 0.088 0.105 15.654 15.733 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 12.622 12.953 init_scf_run 11 5.9 0.000 0.001 12.186 12.188 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.186 12.187 make_images_data 5054 15.6 0.063 0.071 9.385 11.443 density_rs2pw 129 9.7 0.006 0.007 8.843 11.206 multiply_cannon_metrocomm4 27797 15.6 0.100 0.115 3.796 11.026 hybrid_alltoall_any 5240 16.5 0.345 1.523 7.898 10.712 mp_irecv_dv 70031 16.3 3.599 10.628 3.599 10.628 rs_pw_transfer 1054 12.0 0.014 0.016 7.928 10.383 dbcsr_mm_accdrv_process 62734 16.2 4.522 5.213 9.105 9.788 pw_transfer 1559 11.6 0.087 0.100 9.449 9.531 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 9.221 9.307 wfi_extrapolate 11 7.9 0.001 0.001 8.800 8.800 cp_fm_cholesky_decompose 22 10.9 8.464 8.541 8.464 8.541 fft_wrap_pw1pw2_140 527 13.2 1.338 1.360 7.794 7.903 grid_integrate_task_list 129 12.3 7.218 7.812 7.218 7.812 mp_sum_l 7930 13.1 5.632 7.802 5.632 7.802 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.729 7.468 calculate_dm_sparse 129 9.5 0.001 0.001 6.687 6.841 fft3d_ps 1301 14.7 2.838 3.010 6.519 6.589 potential_pw2rs 129 12.3 0.015 0.018 6.354 6.397 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.294 6.302 grid_collocate_task_list 129 9.7 4.740 5.902 4.740 5.902 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.795 5.876 mp_allgather_i34 2527 14.6 2.499 5.863 2.499 5.863 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.432 5.581 mp_alltoall_d11v 2423 14.1 4.521 5.547 4.521 5.547 mp_waitany 11836 13.9 2.666 5.145 2.666 5.145 rs_pw_transfer_RS2PW_140 140 11.5 0.357 0.385 2.128 4.490 dbcsr_complete_redistribute 395 12.7 0.767 0.855 3.474 4.383 mp_sum_d 4499 12.1 2.942 4.368 2.942 4.368 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=209.561000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=792.636364, yerr=5.580367 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 941.293568E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172558800 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58198. MP_Allreduce 11005 1007. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.030 195.194 195.195 qs_mol_dyn_low 1 2.0 0.003 0.003 194.640 194.654 qs_forces 11 3.9 0.004 0.005 194.508 194.516 qs_energies 11 4.9 0.003 0.013 187.653 187.676 scf_env_do_scf 11 5.9 0.001 0.001 170.861 170.862 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 131.242 131.243 velocity_verlet 10 3.0 0.001 0.001 122.784 122.787 qs_scf_new_mos 116 7.6 0.001 0.001 90.463 90.806 qs_scf_loop_do_ot 116 8.6 0.001 0.001 90.462 90.805 dbcsr_multiply_generic 2485 12.5 0.207 0.214 85.614 87.463 ot_scf_mini 116 9.6 0.003 0.004 86.225 86.632 multiply_cannon 2485 13.5 0.499 0.517 62.934 68.102 multiply_cannon_loop 2485 14.5 0.847 0.873 59.287 64.084 ot_mini 116 10.6 0.001 0.001 45.628 46.036 mp_waitall_1 169034 16.6 27.983 40.086 27.983 40.086 init_scf_loop 11 6.9 0.000 0.000 39.508 39.510 prepare_preconditioner 11 7.9 0.000 0.000 35.305 35.352 make_preconditioner 11 8.9 0.000 0.000 35.305 35.352 make_full_inverse_cholesky 11 9.9 0.000 0.000 32.918 34.332 rebuild_ks_matrix 127 8.3 0.001 0.001 32.236 32.941 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 32.236 32.941 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.001 29.658 multiply_cannon_metrocomm3 19880 15.5 0.058 0.064 16.136 27.829 qs_ot_get_derivative 116 11.6 0.001 0.002 25.705 26.107 qs_ot_get_p 127 10.4 0.001 0.001 25.437 25.921 multiply_cannon_multrec 19880 15.5 13.274 16.203 21.947 24.804 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.845 20.830 apply_single 127 13.6 0.001 0.001 19.844 20.829 qs_ot_p2m_diag 82 11.4 0.262 0.269 20.402 20.484 ot_diis_step 116 11.6 0.018 0.018 19.765 19.766 cp_dbcsr_syevd 82 12.4 0.005 0.005 19.124 19.126 make_m2s 4970 13.5 0.080 0.084 16.229 17.388 cp_fm_cholesky_invert 11 10.9 17.364 17.378 17.364 17.378 make_images 4970 14.5 1.165 1.259 15.998 17.154 sum_up_and_integrate 127 10.3 0.132 0.143 15.805 15.836 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.730 15.781 calculate_rho_elec 127 8.7 0.129 0.144 15.729 15.781 multiply_cannon_sync_h2d 19880 15.5 14.022 15.745 14.022 15.745 integrate_v_rspace 127 11.3 0.004 0.004 15.672 15.697 cp_fm_diag_elpa 82 13.4 0.000 0.000 15.204 15.223 cp_fm_diag_elpa_base 82 14.4 14.714 14.925 15.199 15.219 make_images_data 4970 15.5 0.058 0.066 10.137 11.698 init_scf_run 11 5.9 0.000 0.001 11.474 11.474 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.474 11.474 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.610 10.890 hybrid_alltoall_any 5155 16.4 0.428 1.964 8.602 10.731 density_rs2pw 127 9.7 0.006 0.006 8.475 10.255 pw_transfer 1535 11.6 0.086 0.104 9.502 9.623 multiply_cannon_metrocomm4 17395 15.5 0.063 0.075 3.442 9.558 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 9.279 9.408 mp_irecv_dv 49801 16.2 3.321 9.310 3.321 9.310 cp_fm_cholesky_decompose 22 10.9 8.802 8.819 8.802 8.819 rs_pw_transfer 1038 11.9 0.013 0.014 7.005 8.762 dbcsr_mm_accdrv_process 41158 16.2 4.417 4.953 8.152 8.257 fft_wrap_pw1pw2_140 519 13.2 1.400 1.421 7.906 8.039 wfi_extrapolate 11 7.9 0.001 0.001 8.011 8.011 grid_integrate_task_list 127 12.3 7.174 7.846 7.174 7.846 cp_fm_upper_to_full 104 14.8 5.748 7.337 5.748 7.337 dbcsr_complete_redistribute 393 12.7 1.251 1.301 5.002 6.794 fft3d_ps 1281 14.7 2.722 2.952 6.428 6.537 mp_sum_l 7804 13.0 4.625 6.489 4.625 6.489 calculate_dm_sparse 127 9.5 0.001 0.001 5.890 5.986 potential_pw2rs 127 12.3 0.020 0.022 5.822 5.874 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.789 5.797 mp_alltoall_d11v 2401 14.1 4.966 5.753 4.966 5.753 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.015 5.705 grid_collocate_task_list 127 9.7 4.872 5.664 4.872 5.664 mp_allgather_i34 2485 14.5 2.171 5.478 2.171 5.478 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.661 5.453 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.681 4.885 mp_waitany 11660 13.9 2.511 4.447 2.511 4.447 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.391 4.436 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.367 4.136 mp_alltoall_i22 712 14.1 2.120 4.094 2.120 4.094 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=195.195000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=892.181818, yerr=10.928014 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.121448E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669921808880 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57632. MP_Allreduce 11104 1087. MP_Sync 88 MP_Alltoall 1724 12509633. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.075 0.127 208.927 208.929 qs_mol_dyn_low 1 2.0 0.003 0.004 208.247 208.259 qs_forces 11 3.9 0.004 0.004 208.137 208.144 qs_energies 11 4.9 0.008 0.035 200.799 200.816 scf_env_do_scf 11 5.9 0.001 0.003 183.045 183.052 velocity_verlet 10 3.0 0.001 0.001 136.566 136.569 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 131.094 131.096 qs_scf_new_mos 118 7.6 0.001 0.001 90.725 91.057 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.725 91.056 ot_scf_mini 118 9.6 0.003 0.004 86.098 86.426 dbcsr_multiply_generic 2529 12.6 0.190 0.201 83.062 84.102 multiply_cannon 2529 13.6 0.558 0.604 56.268 58.917 multiply_cannon_loop 2529 14.6 1.189 1.220 51.994 54.539 init_scf_loop 11 6.9 0.000 0.002 51.808 51.812 prepare_preconditioner 11 7.9 0.000 0.001 47.493 47.527 make_preconditioner 11 8.9 0.000 0.001 47.493 47.527 make_full_inverse_cholesky 11 9.9 0.000 0.001 41.264 46.245 ot_mini 118 10.6 0.001 0.001 45.358 45.691 mp_waitall_1 149172 16.7 19.992 32.031 19.992 32.031 rebuild_ks_matrix 129 8.3 0.001 0.001 31.726 31.992 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.024 31.726 31.991 multiply_cannon_multrec 30348 15.6 14.036 19.462 26.043 30.952 qs_ks_update_qs_env 129 7.6 0.001 0.001 28.663 28.918 qs_ot_get_p 129 10.4 0.001 0.001 26.469 26.801 qs_ot_get_derivative 118 11.6 0.002 0.002 25.218 25.547 make_m2s 5058 13.6 0.098 0.102 21.895 23.527 make_images 5058 14.6 1.959 2.276 21.582 23.218 qs_ot_p2m_diag 84 11.4 0.348 0.395 21.756 21.829 cp_dbcsr_syevd 84 12.4 0.006 0.007 20.232 20.234 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.468 20.117 apply_single 129 13.6 0.001 0.001 19.467 20.117 ot_diis_step 118 11.6 0.018 0.026 20.005 20.006 cp_fm_cholesky_invert 11 10.9 19.799 19.811 19.799 19.811 multiply_cannon_metrocomm3 30348 15.6 0.048 0.050 7.066 17.141 cp_fm_upper_to_full 106 14.8 11.655 17.067 11.655 17.067 cp_fm_diag_elpa 84 13.4 0.001 0.001 16.357 16.372 cp_fm_diag_elpa_base 84 14.4 15.290 15.630 16.348 16.363 sum_up_and_integrate 129 10.3 0.141 0.152 16.204 16.232 integrate_v_rspace 129 11.3 0.004 0.004 16.063 16.096 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.006 16.052 calculate_rho_elec 129 8.7 0.175 0.191 16.005 16.052 make_images_data 5058 15.6 0.063 0.069 11.998 14.231 hybrid_alltoall_any 5245 16.5 0.528 2.207 10.659 13.015 dbcsr_complete_redistribute 397 12.7 1.511 1.614 9.211 13.001 multiply_cannon_sync_h2d 30348 15.6 11.770 12.503 11.770 12.503 dbcsr_mm_accdrv_process 62780 16.2 7.469 8.456 11.580 12.060 init_scf_run 11 5.9 0.000 0.002 11.644 11.645 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.644 11.645 copy_fm_to_dbcsr 210 11.7 0.002 0.002 7.704 11.520 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 10.795 11.028 pw_transfer 1559 11.6 0.087 0.098 9.994 10.105 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.206 9.965 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.765 9.878 density_rs2pw 129 9.7 0.006 0.008 8.268 9.781 cp_fm_cholesky_decompose 22 10.9 9.202 9.309 9.202 9.309 mp_alltoall_i22 720 14.1 5.510 9.286 5.510 9.286 fft_wrap_pw1pw2_140 527 13.2 1.564 1.596 8.454 8.570 wfi_extrapolate 11 7.9 0.001 0.001 8.169 8.169 grid_integrate_task_list 129 12.3 7.560 8.126 7.560 8.126 rs_pw_transfer 1054 12.0 0.013 0.015 6.456 8.073 multiply_cannon_metrocomm4 25290 15.6 0.084 0.097 2.839 7.698 mp_irecv_dv 76751 16.2 2.686 7.412 2.686 7.412 mp_alltoall_d11v 2429 14.1 5.595 7.001 5.595 7.001 fft3d_ps 1301 14.7 2.864 2.938 6.668 6.746 calculate_dm_sparse 129 9.5 0.001 0.001 6.350 6.485 grid_collocate_task_list 129 9.7 5.110 5.780 5.110 5.780 potential_pw2rs 129 12.3 0.024 0.024 5.714 5.760 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.543 5.591 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.762 4.855 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.522 4.583 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.416 4.505 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.499 4.500 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=208.929000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1064.545455, yerr=14.118741 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 5.910121E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.516249E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+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 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592338480 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57630. MP_Allreduce 11104 1171. MP_Sync 88 MP_Alltoall 1724 18848050. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.070 199.578 199.594 qs_mol_dyn_low 1 2.0 0.003 0.003 199.030 199.041 qs_forces 11 3.9 0.004 0.005 198.916 198.921 qs_energies 11 4.9 0.001 0.001 191.122 191.132 scf_env_do_scf 11 5.9 0.001 0.001 172.657 172.668 scf_env_do_scf_inner_loop 118 6.6 0.003 0.010 130.659 130.660 velocity_verlet 10 3.0 0.001 0.001 128.611 128.616 qs_scf_new_mos 118 7.6 0.001 0.001 88.021 88.142 qs_scf_loop_do_ot 118 8.6 0.001 0.001 88.020 88.141 ot_scf_mini 118 9.6 0.004 0.004 83.323 83.368 dbcsr_multiply_generic 2529 12.6 0.183 0.194 79.314 79.950 multiply_cannon 2529 13.6 0.591 0.629 57.063 62.667 multiply_cannon_loop 2529 14.6 0.446 0.458 51.653 53.171 ot_mini 118 10.6 0.001 0.001 43.166 43.252 init_scf_loop 11 6.9 0.000 0.000 41.832 41.834 mp_waitall_1 126876 16.7 29.346 38.249 29.346 38.249 prepare_preconditioner 11 7.9 0.000 0.000 37.717 37.743 make_preconditioner 11 8.9 0.000 0.000 37.717 37.743 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.439 35.816 rebuild_ks_matrix 129 8.3 0.001 0.001 32.069 32.164 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.019 32.069 32.163 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.201 29.278 qs_ot_get_p 129 10.4 0.001 0.001 26.580 26.705 qs_ot_get_derivative 118 11.6 0.002 0.002 22.593 22.643 cp_fm_cholesky_invert 11 10.9 22.484 22.494 22.484 22.494 qs_ot_p2m_diag 84 11.4 0.502 0.507 22.117 22.132 multiply_cannon_multrec 10116 15.6 10.496 14.573 17.829 21.180 cp_dbcsr_syevd 84 12.4 0.005 0.006 20.722 20.724 make_m2s 5058 13.6 0.068 0.073 18.017 20.624 multiply_cannon_metrocomm3 10116 15.6 0.023 0.024 13.045 20.589 ot_diis_step 118 11.6 0.020 0.023 20.446 20.446 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.035 20.371 apply_single 129 13.6 0.001 0.001 20.035 20.371 make_images 5058 14.6 2.299 2.752 17.706 20.319 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.055 17.077 calculate_rho_elec 129 8.7 0.258 0.269 17.054 17.076 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.811 16.821 cp_fm_diag_elpa_base 84 14.4 16.487 16.637 16.806 16.815 sum_up_and_integrate 129 10.3 0.181 0.190 16.620 16.666 integrate_v_rspace 129 11.3 0.004 0.004 16.439 16.492 make_images_data 5058 15.6 0.051 0.061 11.008 14.215 hybrid_alltoall_any 5245 16.5 0.849 3.839 10.605 13.955 multiply_cannon_sync_h2d 10116 15.6 11.686 12.344 11.686 12.344 init_scf_run 11 5.9 0.000 0.001 11.457 11.457 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.456 11.457 pw_transfer 1559 11.6 0.087 0.097 10.596 10.649 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 10.370 10.431 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.494 9.541 cp_fm_cholesky_decompose 22 10.9 9.238 9.347 9.238 9.347 density_rs2pw 129 9.7 0.006 0.006 8.264 9.135 fft_wrap_pw1pw2_140 527 13.2 1.916 1.949 8.933 8.993 mp_allgather_i34 2529 14.6 3.593 8.413 3.593 8.413 multiply_cannon_metrocomm1 10116 15.6 0.029 0.029 5.411 8.199 grid_integrate_task_list 129 12.3 7.782 8.145 7.782 8.145 wfi_extrapolate 11 7.9 0.001 0.001 8.075 8.075 dbcsr_mm_accdrv_process 20934 16.1 2.687 3.481 6.967 7.597 mp_alltoall_d11v 2429 14.1 6.298 7.121 6.298 7.121 fft3d_ps 1301 14.7 2.787 2.862 6.835 6.892 rs_pw_transfer 1054 12.0 0.012 0.014 5.778 6.747 calculate_dm_sparse 129 9.5 0.001 0.001 6.425 6.506 dbcsr_complete_redistribute 397 12.7 2.116 2.150 5.748 6.175 grid_collocate_task_list 129 9.7 5.440 6.110 5.440 6.110 potential_pw2rs 129 12.3 0.026 0.027 5.576 5.622 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.366 5.371 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.342 5.343 multiply_cannon_metrocomm4 7587 15.6 0.024 0.028 1.864 4.323 mp_irecv_dv 29102 15.9 1.830 4.255 1.830 4.255 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.199 4.250 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.151 4.218 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.733 4.105 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.889 3.997 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=199.594000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1396.090909, yerr=50.683428 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1979979923456 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.250208E+12 0.0% 0.0% 100.0% flops max/rank 11.788211E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806645728 0.0% 0.0% 100.0% number of processed stacks 1983232 0.0% 0.0% 100.0% average stack size 0.0 0.0 3432.1 marketing flops 145.668111E+12 ------------------------------------------------------------------------------- # multiplications 2537 max memory usage/rank 2.963870E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101480 MPI messages size (bytes): total size 1.145238E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.285357E+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 45968 35701915648 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4056 58383. MP_Allreduce 11137 1510. MP_Sync 87 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.045 0.169 317.562 317.649 qs_mol_dyn_low 1 2.0 0.003 0.003 316.900 316.908 qs_forces 11 3.9 0.005 0.006 316.668 316.672 qs_energies 11 4.9 0.001 0.001 307.086 307.099 scf_env_do_scf 11 5.9 0.001 0.001 283.662 283.683 velocity_verlet 10 3.0 0.001 0.001 227.339 227.362 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 152.660 152.662 init_scf_loop 11 6.9 0.000 0.000 130.711 130.713 prepare_preconditioner 11 7.9 0.000 0.000 125.370 125.399 make_preconditioner 11 8.9 0.000 0.000 125.370 125.399 make_full_inverse_cholesky 11 9.9 0.000 0.000 100.600 122.433 qs_scf_new_mos 118 7.6 0.001 0.001 97.937 98.068 qs_scf_loop_do_ot 118 8.6 0.001 0.001 97.936 98.067 ot_scf_mini 118 9.6 0.004 0.004 93.086 93.143 dbcsr_multiply_generic 2537 12.6 0.212 0.225 86.249 86.844 cp_fm_upper_to_full 105 14.8 53.694 77.518 53.694 77.518 multiply_cannon 2537 13.6 0.711 0.772 59.384 60.572 multiply_cannon_loop 2537 14.6 0.473 0.479 55.312 56.688 ot_mini 118 10.6 0.001 0.001 46.890 46.951 dbcsr_complete_redistribute 395 12.7 4.025 4.091 30.904 45.009 copy_fm_to_dbcsr 209 11.7 0.002 0.002 27.265 41.367 rebuild_ks_matrix 129 8.3 0.001 0.001 40.872 40.915 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 40.871 40.914 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.725 38.664 qs_ks_update_qs_env 129 7.6 0.001 0.001 37.654 37.694 mp_alltoall_i22 718 14.1 22.498 36.854 22.498 36.854 cp_fm_cholesky_invert 11 10.9 36.237 36.243 36.237 36.243 mp_waitall_1 104900 16.8 30.306 34.702 30.306 34.702 qs_ot_get_p 129 10.4 0.001 0.001 30.317 30.377 qs_ot_p2m_diag 83 11.4 0.878 0.883 25.870 25.890 qs_ot_get_derivative 118 11.6 0.002 0.002 25.799 25.858 cp_dbcsr_syevd 83 12.4 0.006 0.006 24.066 24.068 make_m2s 5074 13.6 0.078 0.081 22.190 23.400 make_images 5074 14.6 3.783 4.038 21.706 22.919 qs_rho_update_rho_low 129 7.7 0.001 0.001 22.532 22.549 calculate_rho_elec 129 8.7 0.484 0.485 22.532 22.548 sum_up_and_integrate 129 10.3 0.324 0.325 21.485 21.568 integrate_v_rspace 129 11.3 0.004 0.004 21.161 21.244 ot_diis_step 118 11.6 0.022 0.023 21.053 21.054 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.239 20.456 apply_single 129 13.6 0.001 0.001 20.238 20.456 cp_fm_diag_elpa 83 13.4 0.000 0.000 20.358 20.358 cp_fm_diag_elpa_base 83 14.4 15.760 17.591 20.352 20.353 multiply_cannon_metrocomm3 10148 15.6 0.023 0.024 18.730 20.036 multiply_cannon_multrec 10148 15.6 10.584 12.432 18.057 18.114 multiply_cannon_sync_h2d 10148 15.6 15.785 15.795 15.785 15.795 make_images_data 5074 15.6 0.060 0.065 12.325 14.718 hybrid_alltoall_any 5260 16.5 1.308 3.054 12.240 14.619 pw_transfer 1559 11.6 0.094 0.094 13.418 13.423 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 13.181 13.187 init_scf_run 11 5.9 0.000 0.001 13.025 13.026 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.024 13.026 fft_wrap_pw1pw2_140 527 13.2 3.097 3.135 11.516 11.526 mp_alltoall_d11v 2423 14.1 10.071 10.972 10.071 10.972 cp_fm_cholesky_decompose 22 10.9 10.587 10.696 10.587 10.696 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.063 10.109 wfi_extrapolate 11 7.9 0.001 0.001 9.649 9.650 density_rs2pw 129 9.7 0.006 0.006 9.426 9.568 dbcsr_mm_accdrv_process 20966 16.1 3.883 5.746 7.232 9.092 grid_integrate_task_list 129 12.3 8.657 8.871 8.657 8.871 fft3d_ps 1301 14.7 2.868 2.876 8.221 8.242 qs_energies_init_hamiltonians 11 5.9 0.031 0.042 8.182 8.182 rs_scatter_matrices 140 9.7 3.732 4.649 6.657 6.881 calculate_dm_sparse 129 9.5 0.001 0.001 6.754 6.840 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.513 6.620 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.379 6.491 grid_collocate_task_list 129 9.7 6.422 6.472 6.422 6.472 rs_gather_matrices 129 12.3 2.554 3.184 6.246 6.457 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=317.649000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2647.181818, yerr=166.645415 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 1.261429E+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 41216215. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.055 85.285 85.286 qs_energies 1 2.0 0.000 0.000 84.775 84.781 ls_scf 1 3.0 0.000 0.000 83.835 83.842 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.597 72.746 multiply_cannon 111 7.7 0.017 0.020 55.956 57.332 multiply_cannon_loop 111 8.7 0.207 0.221 52.543 53.843 ls_scf_main 1 4.0 0.000 0.000 52.505 52.517 density_matrix_trs4 2 5.0 0.002 0.003 46.924 47.045 ls_scf_init_scf 1 4.0 0.000 0.000 28.267 28.268 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.166 27.218 mp_waitall_1 11031 10.9 22.624 25.792 22.624 25.792 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.022 25.050 multiply_cannon_multrec 2664 9.7 8.158 8.908 15.448 17.083 multiply_cannon_sync_h2d 2664 9.7 13.547 15.902 13.547 15.902 make_m2s 222 7.7 0.008 0.011 13.043 13.593 make_images 222 8.7 0.099 0.108 13.021 13.574 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.898 12.692 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.444 8.658 make_images_data 222 9.7 0.004 0.005 7.588 8.121 dbcsr_mm_accdrv_process 4760 10.4 0.506 0.601 6.909 7.863 hybrid_alltoall_any 227 10.6 0.215 1.826 6.465 7.823 dbcsr_mm_accdrv_process_sort 4760 11.4 6.203 7.101 6.203 7.101 calculate_norms 4752 9.8 5.509 6.099 5.509 6.099 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.075 5.232 mp_sum_l 807 5.4 3.072 4.346 3.072 4.346 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.053 3.770 mp_irecv_dv 6231 10.9 2.036 3.741 2.036 3.741 arnoldi_extremal 4 6.8 0.000 0.000 3.446 3.480 arnoldi_normal_ev 4 7.8 0.001 0.003 3.446 3.480 build_subspace 16 8.4 0.009 0.012 3.335 3.340 make_images_sizes 222 9.7 0.000 0.000 0.780 3.309 mp_alltoall_i44 222 10.7 0.779 3.309 0.779 3.309 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.275 3.109 ls_scf_post 1 4.0 0.000 0.000 3.063 3.070 ls_scf_store_result 1 5.0 0.000 0.000 2.888 2.925 dbcsr_special_finalize 555 9.7 0.005 0.006 2.327 2.754 dbcsr_merge_single_wm 555 10.7 0.450 0.583 2.319 2.744 make_images_pack 222 9.7 2.202 2.644 2.204 2.645 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.377 2.619 dbcsr_sort_data 658 11.4 2.126 2.486 2.126 2.486 dbcsr_matrix_vector_mult_local 304 10.0 2.058 2.465 2.060 2.467 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.320 2.419 buffer_matrices_ensure_size 222 8.7 1.742 2.040 1.742 2.040 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.804 1.805 rebuild_ks_matrix 3 7.3 0.000 0.000 1.794 1.796 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.794 1.796 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.286000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1130.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 2.133037E+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 0.030 0.053 91.281 91.283 qs_energies 1 2.0 0.000 0.000 90.803 90.807 ls_scf 1 3.0 0.000 0.000 89.486 89.490 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.470 75.763 multiply_cannon 111 7.7 0.029 0.042 53.573 57.520 ls_scf_main 1 4.0 0.000 0.000 55.069 55.074 multiply_cannon_loop 111 8.7 0.115 0.122 50.264 53.404 density_matrix_trs4 2 5.0 0.002 0.003 49.323 49.525 ls_scf_init_scf 1 4.0 0.000 0.000 30.891 30.892 mp_waitall_1 9105 10.9 21.254 30.588 21.254 30.588 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.720 29.801 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.333 27.347 multiply_cannon_multrec 1332 9.7 13.172 16.355 22.365 26.371 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 12.017 21.256 make_m2s 222 7.7 0.006 0.008 15.157 15.847 make_images 222 8.7 1.567 1.957 15.127 15.818 dbcsr_mm_accdrv_process 4041 10.4 0.306 0.458 8.789 10.367 dbcsr_mm_accdrv_process_sort 4041 11.4 8.381 9.912 8.381 9.912 make_images_data 222 9.7 0.004 0.004 8.629 9.802 hybrid_alltoall_any 227 10.6 0.522 2.479 8.069 9.071 mp_sum_l 807 5.4 5.444 8.476 5.444 8.476 multiply_cannon_metrocomm4 1221 9.7 0.006 0.009 3.232 8.030 mp_irecv_dv 3311 11.0 3.212 7.984 3.212 7.984 calculate_norms 2376 9.8 5.994 6.746 5.994 6.746 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.240 6.735 multiply_cannon_sync_h2d 1332 9.7 4.820 6.194 4.820 6.194 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.063 5.270 arnoldi_extremal 4 6.8 0.000 0.000 4.747 4.769 arnoldi_normal_ev 4 7.8 0.001 0.005 4.747 4.769 build_subspace 16 8.4 0.014 0.021 4.483 4.487 ls_scf_post 1 4.0 0.000 0.000 3.527 3.532 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.181 3.404 ls_scf_store_result 1 5.0 0.000 0.000 3.233 3.353 dbcsr_matrix_vector_mult_local 304 10.0 2.736 3.218 2.738 3.220 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.600 2.661 mp_allgather_i34 111 8.7 1.007 2.445 1.007 2.445 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.222 2.432 make_images_pack 222 9.7 2.020 2.414 2.022 2.416 dbcsr_sort_data 436 11.2 1.830 2.049 1.830 2.049 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.880 1.882 rebuild_ks_matrix 3 7.3 0.000 0.000 1.867 1.869 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.867 1.869 dbcsr_data_new 4174 10.1 1.609 1.842 1.609 1.842 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.283000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1750.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 2.802815E+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 0.031 0.050 94.897 94.898 qs_energies 1 2.0 0.000 0.000 94.246 94.249 ls_scf 1 3.0 0.000 0.000 92.850 92.854 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.223 77.572 ls_scf_main 1 4.0 0.000 0.000 58.176 58.182 multiply_cannon 111 7.7 0.039 0.107 53.097 57.067 multiply_cannon_loop 111 8.7 0.100 0.110 49.452 54.023 density_matrix_trs4 2 5.0 0.002 0.003 52.096 52.286 mp_waitall_1 7281 11.0 23.852 34.135 23.852 34.135 ls_scf_init_scf 1 4.0 0.000 0.000 31.077 31.078 ls_scf_init_matrix_S 1 5.0 0.000 0.002 29.879 29.947 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.495 27.512 multiply_cannon_multrec 888 9.7 12.657 15.674 21.147 24.255 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.140 23.474 make_m2s 222 7.7 0.006 0.007 16.940 18.218 make_images 222 8.7 1.970 2.285 16.901 18.177 make_images_data 222 9.7 0.003 0.004 9.528 10.660 hybrid_alltoall_any 227 10.6 0.622 2.869 9.076 10.642 mp_sum_l 807 5.4 5.626 9.332 5.626 9.332 dbcsr_mm_accdrv_process 3754 10.4 0.241 0.418 8.027 9.256 dbcsr_mm_accdrv_process_sort 3754 11.4 7.657 8.837 7.657 8.837 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.403 7.590 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.781 7.431 multiply_cannon_sync_h2d 888 9.7 6.016 7.355 6.016 7.355 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.423 6.813 mp_irecv_dv 2335 11.1 2.409 6.775 2.409 6.775 arnoldi_extremal 4 6.8 0.000 0.000 5.392 5.416 arnoldi_normal_ev 4 7.8 0.001 0.004 5.392 5.416 apply_matrix_preconditioner 6 5.3 0.000 0.001 4.990 5.175 build_subspace 16 8.4 0.014 0.020 5.082 5.088 calculate_norms 1584 9.8 4.293 4.754 4.293 4.754 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.600 3.874 mp_allgather_i34 111 8.7 1.481 3.824 1.481 3.824 ls_scf_post 1 4.0 0.000 0.000 3.596 3.600 dbcsr_matrix_vector_mult_local 304 10.0 3.008 3.583 3.010 3.585 ls_scf_store_result 1 5.0 0.000 0.000 3.314 3.428 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.876 2.954 make_images_sizes 222 9.7 0.000 0.000 1.147 2.336 mp_alltoall_i44 222 10.7 1.147 2.336 1.147 2.336 dbcsr_sort_data 325 11.1 1.851 2.147 1.851 2.147 make_images_pack 222 9.7 1.811 2.098 1.814 2.101 dbcsr_data_release 9322 10.9 1.289 1.983 1.289 1.983 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.952 1.953 rebuild_ks_matrix 3 7.3 0.000 0.000 1.933 1.935 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.933 1.935 dbcsr_finalize 304 7.8 0.026 0.032 1.614 1.917 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.898000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2208.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 3.371483E+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 0.074 0.144 96.995 97.001 qs_energies 1 2.0 0.000 0.000 95.972 95.989 ls_scf 1 3.0 0.000 0.001 94.292 94.309 dbcsr_multiply_generic 111 6.7 0.017 0.017 77.581 77.839 ls_scf_main 1 4.0 0.000 0.001 58.340 58.343 multiply_cannon 111 7.7 0.058 0.163 51.151 55.742 density_matrix_trs4 2 5.0 0.004 0.048 52.168 52.305 multiply_cannon_loop 111 8.7 0.115 0.126 46.008 48.416 ls_scf_init_scf 1 4.0 0.000 0.002 32.723 32.725 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.483 31.544 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.025 29.039 mp_waitall_1 6369 11.0 22.120 28.333 22.120 28.333 multiply_cannon_multrec 1332 9.7 14.191 17.116 22.011 24.555 make_m2s 222 7.7 0.007 0.008 20.919 22.439 make_images 222 8.7 3.133 3.613 20.869 22.391 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 8.838 16.530 make_images_data 222 9.7 0.004 0.004 11.635 13.232 hybrid_alltoall_any 227 10.6 0.798 3.741 10.931 12.586 dbcsr_mm_accdrv_process 3641 10.4 0.212 0.406 7.462 8.989 dbcsr_mm_accdrv_process_sort 3641 11.4 7.089 8.568 7.089 8.568 mp_sum_l 807 5.4 4.042 7.204 4.042 7.204 multiply_cannon_sync_h2d 1332 9.7 5.450 6.006 5.450 6.006 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.082 5.956 mp_irecv_dv 3229 10.9 2.059 5.879 2.059 5.879 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.021 5.700 arnoldi_extremal 4 6.8 0.000 0.000 5.619 5.641 arnoldi_normal_ev 4 7.8 0.001 0.005 5.619 5.641 build_subspace 16 8.4 0.014 0.021 5.266 5.273 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.468 5.164 apply_matrix_preconditioner 6 5.3 0.000 0.001 4.567 4.728 mp_allgather_i34 111 8.7 2.260 4.664 2.260 4.664 calculate_norms 2376 9.8 4.208 4.535 4.208 4.535 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.743 4.034 dbcsr_matrix_vector_mult_local 304 10.0 3.173 3.669 3.175 3.670 dbcsr_sort_data 658 11.4 3.016 3.266 3.016 3.266 ls_scf_post 1 4.0 0.000 0.001 3.229 3.249 dbcsr_special_finalize 555 9.7 0.006 0.007 2.757 3.163 dbcsr_merge_single_wm 555 10.7 0.533 0.662 2.749 3.155 ls_scf_dm_to_ks 2 5.0 0.001 0.024 3.015 3.079 ls_scf_store_result 1 5.0 0.000 0.000 2.976 3.026 dbcsr_data_release 10477 10.7 1.588 2.367 1.588 2.367 dbcsr_finalize 304 7.8 0.049 0.061 1.801 1.979 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.974 1.976 rebuild_ks_matrix 3 7.3 0.000 0.000 1.951 1.952 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.951 1.952 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.001000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2771.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 4.690997E+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 0.040 0.055 98.976 98.977 qs_energies 1 2.0 0.000 0.000 98.148 98.153 ls_scf 1 3.0 0.000 0.000 96.214 96.225 dbcsr_multiply_generic 111 6.7 0.017 0.019 77.562 77.780 ls_scf_main 1 4.0 0.000 0.000 61.768 61.769 multiply_cannon 111 7.7 0.078 0.146 54.914 60.003 density_matrix_trs4 2 5.0 0.002 0.003 54.637 54.728 multiply_cannon_loop 111 8.7 0.070 0.079 50.404 51.841 mp_waitall_1 5436 11.0 26.197 31.575 26.197 31.575 ls_scf_init_scf 1 4.0 0.000 0.000 30.881 30.886 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.413 29.450 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.250 27.266 multiply_cannon_multrec 444 9.7 14.069 16.533 21.107 24.652 make_m2s 222 7.7 0.004 0.005 17.784 20.284 make_images 222 8.7 3.717 4.413 17.723 20.224 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.112 15.552 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.777 15.435 make_images_data 222 9.7 0.003 0.004 10.110 12.527 hybrid_alltoall_any 227 10.6 0.792 3.791 9.789 12.270 multiply_cannon_sync_h2d 444 9.7 6.552 9.091 6.552 9.091 dbcsr_mm_accdrv_process 3003 10.4 0.209 0.424 6.736 7.857 dbcsr_mm_accdrv_process_sort 3003 11.4 6.399 7.514 6.399 7.514 mp_allgather_i34 111 8.7 2.749 6.992 2.749 6.992 mp_sum_l 807 5.4 2.896 5.906 2.896 5.906 arnoldi_extremal 4 6.8 0.000 0.000 5.805 5.821 arnoldi_normal_ev 4 7.8 0.001 0.005 5.804 5.821 build_subspace 16 8.4 0.015 0.020 5.411 5.423 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.536 4.701 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.055 4.584 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.657 4.503 mp_irecv_dv 1241 11.2 1.638 4.480 1.638 4.480 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.153 4.365 dbcsr_matrix_vector_mult_local 304 10.0 3.678 4.155 3.680 4.158 calculate_norms 792 9.8 3.538 3.701 3.538 3.701 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.527 3.648 make_images_sizes 222 9.7 0.000 0.000 1.037 3.603 mp_alltoall_i44 222 10.7 1.037 3.602 1.037 3.602 ls_scf_post 1 4.0 0.000 0.000 3.565 3.569 ls_scf_store_result 1 5.0 0.000 0.000 3.352 3.392 dbcsr_finalize 304 7.8 0.062 0.078 2.203 2.291 dbcsr_merge_all 275 8.9 0.474 0.526 2.054 2.123 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.073 2.074 rebuild_ks_matrix 3 7.3 0.000 0.000 2.041 2.043 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.041 2.043 dbcsr_data_release 10123 10.8 1.334 1.989 1.334 1.989 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.977000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3639.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 8.768766E+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 0.078 0.097 108.185 108.187 qs_energies 1 2.0 0.000 0.000 106.747 106.763 ls_scf 1 3.0 0.000 0.000 103.820 103.835 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.469 77.584 ls_scf_main 1 4.0 0.000 0.000 65.999 66.000 density_matrix_trs4 2 5.0 0.002 0.003 56.795 56.843 multiply_cannon 111 7.7 0.107 0.193 49.688 51.208 multiply_cannon_loop 111 8.7 0.068 0.074 46.323 47.400 ls_scf_init_scf 1 4.0 0.000 0.000 34.157 34.158 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.619 32.636 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.789 29.803 mp_waitall_1 4527 11.1 22.277 25.733 22.277 25.733 make_m2s 222 7.7 0.005 0.005 24.081 24.974 make_images 222 8.7 4.568 4.953 23.975 24.867 multiply_cannon_multrec 444 9.7 17.881 18.860 22.471 23.233 hybrid_alltoall_any 227 10.6 1.657 3.630 13.063 15.950 make_images_data 222 9.7 0.003 0.003 13.298 15.625 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.393 11.144 multiply_cannon_sync_h2d 444 9.7 8.852 8.899 8.852 8.899 arnoldi_extremal 4 6.8 0.000 0.000 7.392 7.412 arnoldi_normal_ev 4 7.8 0.002 0.009 7.392 7.412 build_subspace 16 8.4 0.026 0.036 6.820 6.833 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.452 5.607 dbcsr_matrix_vector_mult_local 304 10.0 4.998 5.317 5.001 5.320 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.029 5.275 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.158 5.254 dbcsr_mm_accdrv_process 1814 10.4 0.255 0.306 4.419 4.549 dbcsr_mm_accdrv_process_sort 1814 11.4 4.118 4.256 4.118 4.256 ls_scf_post 1 4.0 0.000 0.000 3.665 3.679 make_images_sizes 222 9.7 0.000 0.000 1.467 3.577 mp_alltoall_i44 222 10.7 1.467 3.577 1.467 3.577 ls_scf_store_result 1 5.0 0.000 0.000 3.417 3.427 calculate_norms 792 9.8 3.238 3.269 3.238 3.269 dbcsr_finalize 304 7.8 0.082 0.089 3.084 3.158 mp_allgather_i34 111 8.7 1.015 3.075 1.015 3.075 dbcsr_merge_all 275 8.9 0.890 0.913 2.869 2.939 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.897 2.897 dbcsr_complete_redistribute 5 7.6 1.433 1.459 2.767 2.885 dbcsr_data_release 12724 10.6 2.336 2.858 2.336 2.858 matrix_ls_to_qs 2 6.0 0.000 0.000 2.419 2.544 dbcsr_sort_data 325 11.1 2.438 2.498 2.438 2.498 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.468 2.469 rebuild_ks_matrix 3 7.3 0.000 0.000 2.402 2.404 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.402 2.404 dbcsr_new_transposed 4 7.5 0.243 0.252 2.268 2.277 dbcsr_frobenius_norm 74 6.6 2.057 2.138 2.194 2.237 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.205 dbcsr_add_anytype 103 7.2 0.860 0.891 2.133 2.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.187000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6792.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/97946bc07285c4fbfd3a5916190ff8dec9cb9ad5_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 590.405632E+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 4037197. 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 0.078 0.248 274.365 274.372 qs_mol_dyn_low 1 2.0 0.003 0.005 272.764 272.781 qs_forces 5 3.8 0.005 0.011 272.674 272.676 qs_energies 5 4.8 0.001 0.032 269.057 269.096 scf_env_do_scf 5 5.8 0.001 0.015 252.071 252.074 scf_env_do_scf_inner_loop 105 6.6 0.008 0.051 216.025 216.027 qs_scf_new_mos 105 7.6 0.000 0.001 166.164 166.496 qs_scf_loop_do_ot 105 8.6 0.001 0.002 166.164 166.495 ot_scf_mini 105 9.6 0.003 0.014 156.019 156.310 dbcsr_multiply_generic 1445 12.2 0.144 0.152 138.649 140.005 velocity_verlet 4 3.0 0.001 0.015 121.978 121.980 multiply_cannon 1445 13.2 0.281 0.294 116.893 119.914 multiply_cannon_loop 1445 14.2 2.380 2.502 114.103 117.550 ot_mini 105 10.6 0.001 0.003 63.101 63.229 qs_ot_get_p 112 10.4 0.001 0.001 62.533 62.869 qs_ot_p2m_diag 40 11.0 0.020 0.032 50.228 50.334 mp_waitall_1 488190 16.1 38.729 49.134 38.729 49.134 cp_dbcsr_syevd 40 12.0 0.002 0.002 47.006 47.008 multiply_cannon_multrec 69360 15.2 31.329 37.675 39.599 45.615 qs_ot_get_derivative 55 11.6 0.001 0.003 41.331 41.504 cp_fm_syevd 40 13.0 0.000 0.000 40.835 40.988 multiply_cannon_metrocomm3 69360 15.2 0.192 0.204 26.488 39.177 init_scf_loop 7 6.6 0.001 0.026 35.991 35.994 cp_fm_redistribute_end 40 14.0 17.851 35.600 17.858 35.606 cp_fm_syevd_base 40 14.0 17.732 35.493 17.732 35.493 multiply_cannon_sync_h2d 69360 15.2 30.119 35.059 30.119 35.059 rebuild_ks_matrix 110 8.4 0.000 0.000 33.162 33.388 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.020 33.161 33.388 qs_ks_update_qs_env 112 7.6 0.001 0.001 30.354 30.551 prepare_preconditioner 7 7.6 0.000 0.002 30.215 30.248 make_preconditioner 7 8.6 0.002 0.070 30.215 30.248 qs_rho_update_rho_low 110 7.6 0.001 0.001 23.925 24.356 calculate_rho_elec 110 8.6 0.030 0.033 23.925 24.355 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.098 23.346 apply_single 62 13.6 0.000 0.000 23.097 23.346 make_full_inverse_cholesky 7 9.6 0.000 0.000 21.337 21.401 ot_new_cg_direction 55 11.6 0.001 0.023 21.049 21.051 rs_pw_transfer 690 11.5 0.011 0.012 19.516 20.765 density_rs2pw 110 9.6 0.006 0.007 17.302 18.544 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.214 15.454 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.973 15.084 cp_fm_cholesky_invert 7 10.6 14.527 14.543 14.527 14.543 mp_sum_l 4684 12.4 13.099 14.190 13.099 14.190 init_scf_run 5 5.8 0.000 0.001 13.926 13.927 scf_env_initial_rho_setup 5 6.8 0.002 0.003 13.926 13.927 pw_transfer 1645 12.4 0.086 0.115 13.484 13.810 fft_wrap_pw1pw2 1425 13.5 0.013 0.017 13.339 13.669 calculate_dm_sparse 110 9.5 0.000 0.001 11.789 12.077 sum_up_and_integrate 60 10.3 0.028 0.034 11.851 11.861 integrate_v_rspace 60 11.3 0.002 0.002 11.823 11.833 multiply_cannon_metrocomm1 69360 15.2 0.091 0.097 4.914 11.760 fft_wrap_pw1pw2_240 915 15.0 1.217 1.323 11.146 11.451 qs_vxc_create 110 10.4 0.002 0.004 11.158 11.221 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.877 10.941 check_diag 80 13.5 8.586 8.858 10.654 10.813 fft3d_pb 915 16.0 2.396 2.718 9.173 9.614 make_m2s 2890 13.2 0.077 0.084 8.426 9.133 make_images 2890 14.2 0.243 0.262 8.322 9.028 dbcsr_mm_accdrv_process 154766 15.8 4.009 4.214 8.142 8.920 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 8.715 8.731 make_full_single_inverse 7 9.6 0.001 0.002 8.518 8.552 calculate_first_density_matrix 1 7.0 0.000 0.004 8.296 8.309 acc_transpose_blocks 69360 15.2 0.357 0.374 7.781 8.204 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 8.137 8.192 xc_rho_set_and_dset_create 110 12.4 0.076 0.112 7.647 7.896 potential_pw2rs 60 12.3 0.003 0.003 7.689 7.743 xc_vxc_pw_create 60 11.3 0.038 0.049 7.500 7.563 rs_pw_transfer_RS2PW_30 110 11.6 1.573 1.670 7.128 7.542 mp_sendrecv_dv 168740 12.6 6.768 6.899 6.768 6.899 xc_pw_derive 510 13.4 0.005 0.007 6.766 6.839 mp_alltoall_z22v 2340 17.7 6.217 6.684 6.217 6.684 cp_fm_cholesky_decompose 14 10.2 6.252 6.263 6.252 6.263 acc_transpose_blocks_kernels 69360 16.2 0.848 0.897 5.911 6.216 mp_waitany 7680 13.5 4.696 6.014 4.696 6.014 mp_alltoall_d11v 1300 13.8 4.652 5.901 4.652 5.901 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=274.372000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.800000, yerr=3.655133 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 97946bc07285c4fbfd3a5916190ff8dec9cb9ad5 Summary: empty Status: OK