=== 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: a213519a15191f11b61658d664e7104f22d1fbf8 ################# 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 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, 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 (26.10.2022) # # \ 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 --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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/01 job id: 42700205 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/02 job id: 42700210 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/03 job id: 42700212 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/04 job id: 42700213 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/05 job id: 42700214 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/06 job id: 42700216 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/07 job id: 42700217 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/08 job id: 42700218 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/09 job id: 42700219 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/10 job id: 42700220 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/11 job id: 42700221 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/12 job id: 42700222 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/13 job id: 42700223 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/14 job id: 42700224 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/15 job id: 42700225 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/16 job id: 42700226 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/17 job id: 42700227 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/18 job id: 42700229 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/19 job id: 42700232 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/20 job id: 42700233 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/21 job id: 42700234 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/22 job id: 42700235 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/23 job id: 42700236 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/24 job id: 42700237 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/25 job id: 42700238 --- 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/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/26 job id: 42700239 --- 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 ~~~~~~~ === 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]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 0.021 0.035 132.253 132.254 farming_run 1 2.0 131.769 131.771 132.226 132.230 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.448698E+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 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 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 0.010 0.025 114.889 114.889 qs_energies 1 2.0 0.000 0.000 114.646 114.649 mp2_main 1 3.0 0.000 0.000 112.836 112.839 mp2_gpw_main 1 4.0 0.019 0.026 112.042 112.045 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 93.243 93.689 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.461 55.908 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.168 41.780 47.175 get_2c_integrals 1 6.0 0.000 0.000 36.743 37.609 integrate_v_rspace 273 8.0 0.436 0.449 25.142 30.299 pw_transfer 6555 10.6 0.375 0.382 27.288 27.870 grid_integrate_task_list 273 9.0 20.958 26.603 20.958 26.603 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 25.979 26.570 fft_wrap_pw1pw2_100 2178 12.4 1.149 1.241 23.547 24.147 compute_2c_integrals 1 7.0 0.002 0.002 19.241 19.242 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.793 18.980 mp2_eri_2c_integrate_gpw 1 9.0 2.384 2.427 18.790 18.976 rpa_ri_compute_en 1 5.0 0.000 0.000 18.692 18.893 cp_fm_cholesky_decompose 12 8.2 17.428 18.295 17.428 18.295 cholesky_decomp 1 7.0 0.000 0.000 16.342 17.209 fft3d_s 5443 13.4 16.137 16.635 16.158 16.657 ao_to_mo_and_store_B_mult_1 272 7.0 10.850 15.577 10.850 15.577 calculate_wavefunction 272 8.0 5.428 5.579 12.551 13.175 rpa_num_int 1 6.0 0.000 0.001 10.547 10.548 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.470 10.495 calc_mat_Q 8 8.0 0.000 0.000 9.335 9.443 contract_S_to_Q 8 9.0 0.000 0.000 8.755 8.864 calc_potential_gpw 544 9.5 0.005 0.006 8.261 8.610 parallel_gemm_fm 14 9.1 0.000 0.000 8.334 8.440 parallel_gemm_fm_cosma 14 10.1 8.334 8.440 8.334 8.440 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.169 8.401 potential_pw2rs 545 10.0 0.107 0.109 7.638 8.287 create_integ_mat 1 6.0 0.021 0.028 7.604 7.605 collocate_single_gaussian 272 10.0 0.040 0.043 7.416 7.595 array2fm 1 7.0 0.000 0.000 6.675 7.107 pw_scatter_s 2720 13.7 4.443 4.653 4.443 4.653 pw_gather_s 2722 13.2 3.888 4.275 3.888 4.275 array2fm_buffer_send 1 8.0 2.916 3.146 2.916 3.146 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.045256, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2721.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 0.027 0.036 396.933 396.936 farming_run 1 2.0 396.036 396.040 396.898 396.901 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.222545E+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 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 16305143. MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 209.823 209.824 qs_energies 1 2.0 0.000 0.000 209.485 209.504 scf_env_do_scf 1 3.0 0.000 0.000 106.734 106.735 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.884 105.893 rebuild_ks_matrix 4 6.0 0.000 0.000 105.883 105.892 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.063 105.883 105.892 hfx_ks_matrix 4 8.0 0.001 0.001 105.519 105.523 integrate_four_center 4 9.0 0.144 0.459 105.518 105.522 mp2_main 1 3.0 0.000 0.000 102.473 102.492 mp2_gpw_main 1 4.0 0.033 0.055 101.670 101.689 integrate_four_center_main 4 10.0 0.128 0.602 96.978 99.610 integrate_four_center_bin 270 11.0 96.850 99.567 96.850 99.567 init_scf_loop 1 4.0 0.000 0.000 92.282 92.283 mp2_ri_gpw_compute_in 1 5.0 0.067 0.098 74.878 76.061 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.576 55.756 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.161 42.150 47.622 integrate_v_rspace 95 8.0 0.398 0.567 28.510 33.976 pw_transfer 2240 10.6 0.146 0.171 29.833 30.354 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.864 29.460 grid_integrate_task_list 95 9.0 23.802 29.377 23.802 29.377 ao_to_mo_and_store_B_mult_1 91 7.0 10.728 29.043 10.728 29.043 mp2_ri_gpw_compute_en 1 5.0 0.065 0.083 26.638 28.517 fft_wrap_pw1pw2_100 730 12.4 1.238 1.432 26.623 27.308 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.835 1.889 24.800 24.811 get_2c_integrals 1 6.0 0.001 0.011 20.209 20.239 compute_2c_integrals 1 7.0 0.003 0.006 19.193 19.202 compute_2c_integrals_loop_lm 1 8.0 0.002 0.010 18.746 19.076 mp2_eri_2c_integrate_gpw 1 9.0 1.742 1.844 18.744 19.075 fft3d_s 1823 13.4 18.397 18.881 18.410 18.894 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.451 14.451 calculate_wavefunction 91 8.0 2.019 2.056 9.746 9.983 potential_pw2rs 186 10.0 0.033 0.035 8.582 9.317 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.596 8.745 9.308 local_gemm 172 8.0 8.189 8.750 8.189 8.750 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.011 8.167 8.545 mp2_ri_gpw_compute_en_comm 22 7.0 0.508 0.539 7.831 8.455 collocate_single_gaussian 91 10.0 0.018 0.041 7.829 8.211 calc_potential_gpw 182 9.5 0.002 0.002 7.890 8.123 mp2_ri_gpw_compute_en_ener 172 7.0 6.343 6.418 6.343 6.418 mp_sendrecv_dm3 2068 8.0 5.838 6.410 5.838 6.410 mp_sync 38 10.4 3.216 5.701 3.216 5.701 pw_gather_s 912 13.2 4.950 5.483 4.950 5.483 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.658634, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.211264E+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 271. MP_Sync 580 MP_Alltoall 2083 769776. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.027 50.814 50.815 qs_mol_dyn_low 1 2.0 0.003 0.003 50.618 50.625 qs_forces 11 3.9 0.002 0.002 50.553 50.554 qs_energies 11 4.9 0.001 0.002 49.106 49.116 scf_env_do_scf 11 5.9 0.000 0.001 42.938 42.938 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 40.870 40.870 dbcsr_multiply_generic 2286 12.5 0.092 0.095 32.336 32.762 qs_scf_new_mos 108 7.5 0.000 0.001 30.752 31.053 qs_scf_loop_do_ot 108 8.5 0.000 0.001 30.751 31.052 ot_scf_mini 108 9.5 0.002 0.002 29.030 29.215 multiply_cannon 2286 13.5 0.189 0.207 25.683 27.092 multiply_cannon_loop 2286 14.5 1.482 1.565 24.945 26.402 velocity_verlet 10 3.0 0.001 0.001 25.005 25.006 ot_mini 108 10.5 0.001 0.001 18.527 18.801 qs_ot_get_derivative 108 11.5 0.001 0.001 15.643 15.842 mp_waitall_1 267858 16.1 8.357 14.477 8.357 14.477 multiply_cannon_metrocomm3 54864 15.5 0.069 0.075 5.714 12.707 multiply_cannon_multrec 54864 15.5 4.319 6.743 7.632 11.228 rebuild_ks_matrix 119 8.3 0.000 0.000 7.868 8.039 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.013 7.868 8.039 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.940 7.097 multiply_cannon_sync_h2d 54864 15.5 6.005 6.691 6.005 6.691 mp_sum_l 7207 12.9 4.609 6.415 4.609 6.415 qs_ot_get_p 119 10.4 0.001 0.001 6.068 6.323 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.154 5.599 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.161 5.261 init_scf_run 11 5.9 0.000 0.001 4.955 4.956 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.955 4.955 dbcsr_mm_accdrv_process 76910 16.1 1.117 1.835 3.234 4.579 sum_up_and_integrate 119 10.3 0.012 0.015 4.518 4.525 integrate_v_rspace 119 11.3 0.002 0.002 4.505 4.513 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.243 4.376 calculate_rho_elec 119 8.7 0.011 0.016 4.243 4.375 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.377 3.409 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.897 3.087 apply_single 119 13.6 0.000 0.000 2.897 3.087 calculate_dm_sparse 119 9.5 0.000 0.000 2.949 3.086 rs_pw_transfer 974 11.9 0.011 0.013 2.863 2.950 calculate_first_density_matrix 1 7.0 0.000 0.000 2.795 2.799 jit_kernel_multiply 13 15.8 2.058 2.781 2.058 2.781 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.664 2.664 ot_diis_step 108 11.5 0.006 0.006 2.661 2.661 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.701 2.654 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.452 2.460 density_rs2pw 119 9.7 0.004 0.004 2.349 2.454 cp_fm_redistribute_end 50 14.0 2.223 2.428 2.234 2.441 cp_fm_diag_elpa_base 50 14.0 0.198 2.374 0.199 2.384 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.279 2.332 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.196 2.199 acc_transpose_blocks 54864 15.5 0.217 0.237 1.689 2.156 grid_integrate_task_list 119 12.3 2.023 2.117 2.023 2.117 wfi_extrapolate 11 7.9 0.001 0.001 2.102 2.102 init_scf_loop 11 6.9 0.000 0.000 2.053 2.055 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.947 1.989 potential_pw2rs 119 12.3 0.004 0.004 1.874 1.887 pw_transfer 1439 11.6 0.051 0.054 1.741 1.818 mp_sum_d 4125 12.0 1.209 1.803 1.209 1.803 make_m2s 4572 13.5 0.055 0.057 1.733 1.799 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.667 1.746 make_images 4572 14.5 0.133 0.139 1.649 1.715 fft3d_ps 1201 14.6 0.352 0.453 1.444 1.514 mp_alltoall_d11v 2130 13.8 1.356 1.498 1.356 1.498 mp_waitany 12084 13.8 1.297 1.458 1.297 1.458 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.384 1.406 grid_collocate_task_list 119 9.7 1.285 1.362 1.285 1.362 fft_wrap_pw1pw2_140 487 13.2 0.081 0.093 1.252 1.334 parallel_gemm_fm 81 9.0 0.000 0.000 1.235 1.280 parallel_gemm_fm_cosma 81 10.0 1.235 1.280 1.235 1.280 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.644 1.059 acc_transpose_blocks_kernels 54864 16.5 0.245 0.390 0.746 1.020 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=50.815000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 487.424000E+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 104 MP_Alltoall 2060 1100581. MP_ISendRecv 33558 37093. MP_Wait 40318 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 0.016 0.031 38.945 38.948 qs_mol_dyn_low 1 2.0 0.003 0.003 38.704 38.711 qs_forces 11 3.9 0.002 0.002 38.641 38.648 qs_energies 11 4.9 0.002 0.004 36.947 36.956 scf_env_do_scf 11 5.9 0.000 0.001 31.666 31.666 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.882 28.883 dbcsr_multiply_generic 2286 12.5 0.099 0.102 21.574 21.943 qs_scf_new_mos 108 7.5 0.001 0.001 19.963 20.216 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.963 20.215 ot_scf_mini 108 9.5 0.002 0.003 19.053 19.231 velocity_verlet 10 3.0 0.001 0.001 18.528 18.529 multiply_cannon 2286 13.5 0.210 0.220 16.539 17.997 multiply_cannon_loop 2286 14.5 0.905 0.973 15.421 16.854 ot_mini 108 10.5 0.001 0.001 11.949 12.190 mp_waitall_1 217478 16.2 6.170 11.583 6.170 11.583 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.375 9.770 qs_ot_get_derivative 108 11.5 0.001 0.001 9.470 9.652 multiply_cannon_multrec 27432 15.5 1.965 4.383 5.917 8.753 rebuild_ks_matrix 119 8.3 0.000 0.000 7.162 7.313 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.161 7.312 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.318 6.456 dbcsr_mm_accdrv_process 47894 16.0 3.047 5.224 3.882 5.901 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.720 4.593 qs_ot_get_p 119 10.4 0.001 0.002 4.180 4.411 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.090 4.157 apply_single 119 13.6 0.000 0.000 3.089 4.157 sum_up_and_integrate 119 10.3 0.024 0.027 4.147 4.155 integrate_v_rspace 119 11.3 0.002 0.002 4.122 4.133 mp_sum_l 7207 12.9 2.088 4.094 2.088 4.094 init_scf_run 11 5.9 0.000 0.001 3.834 3.835 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.834 3.835 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.610 3.647 calculate_rho_elec 119 8.7 0.021 0.024 3.609 3.646 multiply_cannon_sync_h2d 27432 15.5 2.197 2.841 2.197 2.841 rs_pw_transfer 974 11.9 0.010 0.011 2.664 2.835 init_scf_loop 11 6.9 0.000 0.002 2.763 2.765 make_m2s 4572 13.5 0.053 0.056 2.507 2.735 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.712 2.730 make_images 4572 14.5 0.200 0.239 2.418 2.644 ot_diis_step 108 11.5 0.011 0.011 2.430 2.431 calculate_first_density_matrix 1 7.0 0.000 0.002 2.325 2.327 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.281 2.282 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.111 2.205 calculate_dm_sparse 119 9.5 0.000 0.001 2.048 2.126 density_rs2pw 119 9.7 0.004 0.004 1.958 2.048 jit_kernel_multiply 10 16.0 0.782 2.014 0.782 2.014 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.952 1.952 cp_fm_redistribute_end 50 14.0 1.606 1.927 1.610 1.929 grid_integrate_task_list 119 12.3 1.830 1.920 1.830 1.920 cp_fm_diag_elpa_base 50 14.0 0.310 1.875 0.318 1.907 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.897 1.900 potential_pw2rs 119 12.3 0.006 0.006 1.883 1.893 pw_transfer 1439 11.6 0.063 0.067 1.790 1.821 prepare_preconditioner 11 7.9 0.000 0.000 1.782 1.813 make_preconditioner 11 8.9 0.000 0.001 1.782 1.813 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.701 1.734 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.687 1.725 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.666 1.720 make_images_data 4572 15.5 0.045 0.051 1.157 1.557 acc_transpose_blocks 27432 15.5 0.107 0.113 1.193 1.491 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.018 1.474 wfi_extrapolate 11 7.9 0.001 0.001 1.458 1.458 fft3d_ps 1201 14.6 0.497 0.550 1.408 1.437 fft_wrap_pw1pw2_140 487 13.2 0.076 0.083 1.333 1.366 mp_alltoall_d11v 2130 13.8 1.259 1.344 1.259 1.344 mp_allgather_i34 2286 14.5 0.560 1.334 0.560 1.334 grid_collocate_task_list 119 9.7 1.239 1.315 1.239 1.315 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.245 1.252 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.199 1.243 qs_energies_init_hamiltonians 11 5.9 0.000 0.003 1.182 1.190 mp_sum_d 4125 12.0 0.642 1.084 0.642 1.084 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.900 0.912 rs_pw_transfer_RS2PW_140 130 11.5 0.161 0.262 0.701 0.869 acc_transpose_blocks_kernels 27432 16.5 0.184 0.275 0.657 0.867 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.608 0.752 0.798 cp_fm_cholesky_invert 11 10.9 0.792 0.796 0.792 0.796 mp_waitany 5720 13.7 0.650 0.794 0.650 0.794 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.948000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.818182, yerr=1.402477 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 520.187904E+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 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 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 0.039 0.064 32.755 32.756 qs_mol_dyn_low 1 2.0 0.003 0.003 32.341 32.348 qs_forces 11 3.9 0.002 0.002 32.261 32.262 qs_energies 11 4.9 0.001 0.002 30.708 30.711 scf_env_do_scf 11 5.9 0.000 0.001 25.613 25.613 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.045 23.045 dbcsr_multiply_generic 2286 12.5 0.152 0.157 16.882 16.975 velocity_verlet 10 3.0 0.001 0.001 15.406 15.408 qs_scf_new_mos 108 7.5 0.001 0.001 14.960 14.980 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.959 14.979 ot_scf_mini 108 9.5 0.002 0.002 14.239 14.259 multiply_cannon 2286 13.5 0.195 0.204 13.258 14.015 multiply_cannon_loop 2286 14.5 0.634 0.663 12.458 13.226 ot_mini 108 10.5 0.001 0.001 8.879 8.902 qs_ot_get_derivative 108 11.5 0.001 0.001 7.320 7.337 multiply_cannon_multrec 18288 15.5 1.945 2.818 6.796 7.014 rebuild_ks_matrix 119 8.3 0.000 0.000 6.384 6.400 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.383 6.399 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.656 5.670 dbcsr_mm_accdrv_process 38222 16.0 4.026 5.380 4.766 5.667 mp_waitall_1 169478 16.3 3.155 4.436 3.155 4.436 sum_up_and_integrate 119 10.3 0.030 0.031 4.008 4.014 integrate_v_rspace 119 11.3 0.002 0.002 3.978 3.987 init_scf_run 11 5.9 0.000 0.001 3.892 3.892 scf_env_initial_rho_setup 11 6.9 0.002 0.004 3.892 3.892 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.745 3.370 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.270 3.276 calculate_rho_elec 119 8.7 0.030 0.031 3.269 3.275 qs_ot_get_p 119 10.4 0.001 0.001 3.148 3.174 calculate_first_density_matrix 1 7.0 0.000 0.000 2.710 2.712 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.429 2.576 init_scf_loop 11 6.9 0.000 0.000 2.550 2.550 rs_pw_transfer 974 11.9 0.009 0.010 2.340 2.424 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.085 2.401 apply_single 119 13.6 0.000 0.000 2.085 2.401 make_m2s 4572 13.5 0.045 0.046 2.056 2.193 make_images 4572 14.5 0.190 0.201 1.971 2.106 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.100 2.101 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.020 2.026 jit_kernel_multiply 10 15.8 0.689 1.963 0.689 1.963 density_rs2pw 119 9.7 0.004 0.004 1.850 1.941 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.891 1.900 grid_integrate_task_list 119 12.3 1.797 1.888 1.797 1.888 calculate_dm_sparse 119 9.5 0.000 0.001 1.808 1.822 potential_pw2rs 119 12.3 0.007 0.008 1.788 1.792 prepare_preconditioner 11 7.9 0.000 0.000 1.778 1.780 make_preconditioner 11 8.9 0.000 0.000 1.778 1.780 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.762 1.763 pw_transfer 1439 11.6 0.063 0.066 1.752 1.762 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.631 1.715 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.662 1.673 multiply_cannon_sync_h2d 18288 15.5 1.406 1.627 1.406 1.627 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.599 1.611 ot_diis_step 108 11.5 0.011 0.011 1.537 1.538 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.531 1.533 cp_fm_redistribute_end 50 14.0 1.136 1.511 1.137 1.511 cp_fm_diag_elpa_base 50 14.0 0.358 1.451 0.372 1.490 mp_sum_l 7207 12.9 1.137 1.451 1.137 1.451 fft3d_ps 1201 14.6 0.504 0.521 1.347 1.358 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.309 1.320 acc_transpose_blocks 18288 15.5 0.074 0.076 1.250 1.296 grid_collocate_task_list 119 9.7 1.204 1.287 1.204 1.287 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.179 1.183 wfi_extrapolate 11 7.9 0.001 0.001 1.136 1.136 make_images_data 4572 15.5 0.045 0.049 0.848 1.061 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.980 0.981 hybrid_alltoall_any 4725 16.4 0.055 0.111 0.729 0.950 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.894 0.918 acc_transpose_blocks_kernels 18288 16.5 0.212 0.222 0.808 0.847 mp_alltoall_d11v 2130 13.8 0.720 0.842 0.720 0.842 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.796 0.798 cp_fm_cholesky_invert 11 10.9 0.743 0.747 0.743 0.747 mp_alltoall_z22v 1201 16.6 0.686 0.746 0.686 0.746 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.658 0.736 multiply_cannon_metrocomm1 18288 15.5 0.028 0.030 0.547 0.715 mp_waitany 9880 13.7 0.579 0.679 0.579 0.679 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.756000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.545455, yerr=1.616035 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 556.851200E+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 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 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 0.016 0.035 34.996 34.997 qs_mol_dyn_low 1 2.0 0.003 0.003 34.763 34.770 qs_forces 11 3.9 0.002 0.002 34.651 34.652 qs_energies 11 4.9 0.002 0.005 32.956 32.961 scf_env_do_scf 11 5.9 0.001 0.002 27.856 27.857 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 24.464 24.464 dbcsr_multiply_generic 2286 12.5 0.098 0.099 18.433 18.537 velocity_verlet 10 3.0 0.001 0.001 17.944 17.948 qs_scf_new_mos 108 7.5 0.001 0.001 16.294 16.349 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.294 16.348 ot_scf_mini 108 9.5 0.002 0.003 15.330 15.385 multiply_cannon 2286 13.5 0.227 0.263 14.640 15.131 multiply_cannon_loop 2286 14.5 0.939 0.971 13.679 14.071 ot_mini 108 10.5 0.001 0.001 9.347 9.410 multiply_cannon_multrec 27432 15.5 2.369 3.060 8.753 9.130 qs_ot_get_derivative 108 11.5 0.001 0.001 7.526 7.580 dbcsr_mm_accdrv_process 47916 15.9 5.242 7.276 6.290 7.523 rebuild_ks_matrix 119 8.3 0.000 0.000 6.462 6.510 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.462 6.509 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.750 5.790 sum_up_and_integrate 119 10.3 0.034 0.037 3.742 3.751 init_scf_run 11 5.9 0.000 0.001 3.750 3.750 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.749 3.750 integrate_v_rspace 119 11.3 0.002 0.003 3.708 3.717 qs_ot_get_p 119 10.4 0.001 0.002 3.315 3.401 init_scf_loop 11 6.9 0.001 0.003 3.371 3.372 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.302 3.336 calculate_rho_elec 119 8.7 0.040 0.046 3.302 3.335 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.742 3.247 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.154 2.618 apply_single 119 13.6 0.000 0.000 2.154 2.618 mp_waitall_1 145218 16.4 2.187 2.582 2.187 2.582 prepare_preconditioner 11 7.9 0.000 0.000 2.541 2.549 make_preconditioner 11 8.9 0.000 0.001 2.541 2.549 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.142 2.472 calculate_first_density_matrix 1 7.0 0.001 0.003 2.408 2.410 make_m2s 4572 13.5 0.055 0.057 2.284 2.404 make_images 4572 14.5 0.271 0.330 2.177 2.295 rs_pw_transfer 974 11.9 0.009 0.009 2.004 2.131 calculate_dm_sparse 119 9.5 0.001 0.001 2.048 2.103 qs_ot_p2m_diag 50 11.0 0.016 0.024 1.967 1.976 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.973 1.974 jit_kernel_multiply 10 15.7 0.988 1.965 0.988 1.965 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.884 1.910 grid_integrate_task_list 119 12.3 1.837 1.906 1.837 1.906 density_rs2pw 119 9.7 0.003 0.004 1.786 1.904 pw_transfer 1439 11.6 0.063 0.066 1.760 1.794 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.769 1.782 ot_diis_step 108 11.5 0.012 0.012 1.780 1.781 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.670 1.707 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.655 1.655 mp_sum_l 7207 12.9 1.091 1.650 1.091 1.650 potential_pw2rs 119 12.3 0.008 0.009 1.523 1.529 acc_transpose_blocks 27432 15.5 0.109 0.111 1.462 1.485 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.419 1.420 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.367 1.409 cp_fm_redistribute_end 50 14.0 0.934 1.397 0.936 1.397 cp_fm_diag_elpa_base 50 14.0 0.439 1.339 0.460 1.380 fft3d_ps 1201 14.6 0.533 0.586 1.351 1.376 grid_collocate_task_list 119 9.7 1.221 1.321 1.221 1.321 wfi_extrapolate 11 7.9 0.001 0.001 1.294 1.294 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.202 1.214 multiply_cannon_metrocomm3 27432 15.5 0.038 0.040 0.732 1.154 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.118 1.132 cp_fm_upper_to_full 72 13.5 0.814 1.129 0.814 1.129 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.088 1.089 dbcsr_complete_redistribute 329 12.2 0.134 0.182 0.795 1.071 multiply_cannon_sync_h2d 27432 15.5 0.999 1.069 0.999 1.069 make_images_data 4572 15.5 0.045 0.048 0.903 1.036 hybrid_alltoall_any 4725 16.4 0.061 0.149 0.775 0.967 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.601 0.872 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.792 0.870 acc_transpose_blocks_kernels 27432 16.5 0.271 0.278 0.834 0.856 mp_alltoall_d11v 2130 13.8 0.728 0.833 0.728 0.833 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.801 0.807 cp_fm_cholesky_invert 11 10.9 0.735 0.738 0.735 0.738 mp_alltoall_i22 627 13.8 0.434 0.735 0.434 0.735 mp_alltoall_z22v 1201 16.6 0.700 0.728 0.700 0.728 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.997000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.181818, yerr=3.352562 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 607.662080E+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 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.038 27.958 27.959 qs_mol_dyn_low 1 2.0 0.003 0.003 27.768 27.775 qs_forces 11 3.9 0.001 0.002 27.711 27.711 qs_energies 11 4.9 0.001 0.002 26.029 26.031 scf_env_do_scf 11 5.9 0.000 0.001 21.325 21.325 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.895 18.896 velocity_verlet 10 3.0 0.001 0.001 14.389 14.392 dbcsr_multiply_generic 2286 12.5 0.091 0.093 12.587 12.677 qs_scf_new_mos 108 7.5 0.001 0.001 11.367 11.386 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.367 11.386 ot_scf_mini 108 9.5 0.002 0.002 10.668 10.695 multiply_cannon 2286 13.5 0.228 0.232 9.962 10.457 multiply_cannon_loop 2286 14.5 0.329 0.340 9.058 9.231 multiply_cannon_multrec 9144 15.5 1.703 1.989 5.951 6.188 ot_mini 108 10.5 0.001 0.001 6.048 6.081 rebuild_ks_matrix 119 8.3 0.000 0.000 5.761 5.784 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.761 5.783 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.141 5.161 qs_ot_get_derivative 108 11.5 0.001 0.001 4.748 4.775 dbcsr_mm_accdrv_process 12550 15.8 3.398 4.222 4.148 4.262 sum_up_and_integrate 119 10.3 0.037 0.040 3.521 3.525 integrate_v_rspace 119 11.3 0.002 0.003 3.484 3.488 init_scf_run 11 5.9 0.000 0.001 3.286 3.287 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.286 3.286 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.193 3.199 calculate_rho_elec 119 8.7 0.059 0.061 3.193 3.198 qs_ot_get_p 119 10.4 0.001 0.001 2.809 2.849 mp_waitall_1 121218 16.5 1.903 2.428 1.903 2.428 init_scf_loop 11 6.9 0.000 0.000 2.410 2.411 calculate_first_density_matrix 1 7.0 0.000 0.000 2.175 2.176 make_m2s 4572 13.5 0.035 0.036 1.822 1.979 grid_integrate_task_list 119 12.3 1.848 1.903 1.848 1.903 make_images 4572 14.5 0.267 0.302 1.732 1.887 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.825 1.827 calculate_dm_sparse 119 9.5 0.000 0.000 1.751 1.768 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.739 1.753 rs_pw_transfer 974 11.9 0.008 0.008 1.632 1.707 prepare_preconditioner 11 7.9 0.000 0.000 1.696 1.700 make_preconditioner 11 8.9 0.000 0.000 1.696 1.700 pw_transfer 1439 11.6 0.063 0.065 1.682 1.692 density_rs2pw 119 9.7 0.003 0.003 1.599 1.681 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.632 1.632 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.619 1.620 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.587 1.614 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.592 1.602 jit_kernel_multiply 8 16.0 0.712 1.462 0.712 1.462 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.441 1.450 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.351 1.352 grid_collocate_task_list 119 9.7 1.271 1.348 1.271 1.348 cp_fm_redistribute_end 50 14.0 0.670 1.329 0.671 1.330 cp_fm_diag_elpa_base 50 14.0 0.613 1.257 0.658 1.315 potential_pw2rs 119 12.3 0.010 0.010 1.305 1.308 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.297 1.306 ot_diis_step 108 11.5 0.013 0.013 1.287 1.287 fft3d_ps 1201 14.6 0.537 0.548 1.260 1.271 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.207 1.208 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.160 1.182 apply_single 119 13.6 0.000 0.000 1.159 1.181 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.145 1.150 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.113 1.127 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.832 1.079 wfi_extrapolate 11 7.9 0.001 0.001 1.068 1.068 make_images_data 4572 15.5 0.039 0.042 0.865 1.053 acc_transpose_blocks 9144 15.5 0.038 0.039 0.939 0.963 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.918 mp_alltoall_d11v 2130 13.8 0.827 0.903 0.827 0.903 cp_fm_cholesky_invert 11 10.9 0.849 0.852 0.849 0.852 multiply_cannon_sync_h2d 9144 15.5 0.702 0.790 0.702 0.790 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.742 0.745 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.727 0.740 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.697 0.718 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.664 0.717 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.354 0.676 mp_allgather_i34 2286 14.5 0.238 0.649 0.238 0.649 mp_alltoall_z22v 1201 16.6 0.596 0.628 0.596 0.628 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.492 0.620 jit_kernel_transpose 5 15.6 0.579 0.600 0.579 0.600 qs_create_task_list 11 7.9 0.000 0.001 0.545 0.569 generate_qs_task_list 11 8.9 0.190 0.213 0.544 0.569 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.959000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=577.454545, yerr=4.250425 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 750.178304E+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_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.032 42.185 42.186 qs_mol_dyn_low 1 2.0 0.003 0.003 41.999 42.005 qs_forces 11 3.9 0.001 0.002 41.924 41.925 qs_energies 11 4.9 0.001 0.001 40.014 40.017 scf_env_do_scf 11 5.9 0.001 0.001 33.067 33.067 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.123 25.123 velocity_verlet 10 3.0 0.001 0.001 24.008 24.016 dbcsr_multiply_generic 2286 12.5 0.098 0.099 18.059 18.213 qs_scf_new_mos 108 7.5 0.001 0.001 16.105 16.196 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.104 16.195 ot_scf_mini 108 9.5 0.002 0.002 15.003 15.100 multiply_cannon 2286 13.5 0.303 0.310 13.970 14.873 multiply_cannon_loop 2286 14.5 0.342 0.347 12.668 13.555 ot_mini 108 10.5 0.001 0.001 9.049 9.164 multiply_cannon_multrec 9144 15.5 3.500 5.186 8.621 8.772 init_scf_loop 11 6.9 0.000 0.000 7.916 7.917 rebuild_ks_matrix 119 8.3 0.000 0.000 7.054 7.193 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.053 7.192 qs_ot_get_derivative 108 11.5 0.001 0.001 7.010 7.110 prepare_preconditioner 11 7.9 0.000 0.000 6.829 6.840 make_preconditioner 11 8.9 0.000 0.000 6.829 6.840 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.384 6.723 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.419 6.546 dbcsr_mm_accdrv_process 12550 15.8 4.129 5.672 4.997 6.390 init_scf_run 11 5.9 0.000 0.001 4.934 4.934 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.934 4.934 cp_fm_upper_to_full 72 14.2 3.177 4.570 3.177 4.570 mp_waitall_1 97218 16.6 3.143 4.104 3.143 4.104 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.904 3.952 calculate_rho_elec 119 8.7 0.118 0.121 3.903 3.951 sum_up_and_integrate 119 10.3 0.064 0.065 3.834 3.841 integrate_v_rspace 119 11.3 0.003 0.003 3.770 3.776 qs_ot_get_p 119 10.4 0.001 0.001 3.130 3.266 calculate_first_density_matrix 1 7.0 0.000 0.000 3.011 3.058 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.607 3.058 dbcsr_complete_redistribute 329 12.2 0.287 0.291 2.179 2.983 make_m2s 4572 13.5 0.038 0.039 2.477 2.693 make_images 4572 14.5 0.349 0.377 2.357 2.573 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.737 2.536 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.226 2.449 apply_single 119 13.6 0.000 0.000 2.226 2.449 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.591 2.421 mp_alltoall_i22 627 13.8 1.525 2.383 1.525 2.383 calculate_dm_sparse 119 9.5 0.000 0.000 2.272 2.310 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.276 2.277 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.440 2.229 pw_transfer 1439 11.6 0.066 0.067 2.220 2.224 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.125 2.129 ot_diis_step 108 11.5 0.014 0.014 2.016 2.016 grid_integrate_task_list 119 12.3 1.992 2.004 1.992 2.004 density_rs2pw 119 9.7 0.003 0.003 1.965 1.988 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.852 1.852 wfi_extrapolate 11 7.9 0.001 0.001 1.810 1.810 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.731 1.779 fft3d_ps 1201 14.6 0.566 0.576 1.760 1.768 mp_sum_l 7207 12.9 1.075 1.762 1.075 1.762 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.746 1.748 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.736 1.741 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.656 1.706 rs_pw_transfer 974 11.9 0.009 0.009 1.566 1.593 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.582 1.582 hybrid_alltoall_any 4725 16.4 0.086 0.146 1.301 1.569 make_images_data 4572 15.5 0.042 0.045 1.252 1.535 grid_collocate_task_list 119 9.7 1.437 1.454 1.437 1.454 mp_alltoall_d11v 2130 13.8 1.345 1.402 1.345 1.402 potential_pw2rs 119 12.3 0.014 0.015 1.385 1.397 cp_fm_cholesky_invert 11 10.9 1.386 1.390 1.386 1.390 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.317 1.360 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.298 1.298 cp_fm_diag_elpa_base 50 14.0 1.152 1.204 1.296 1.296 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.168 1.188 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.083 1.096 mp_alltoall_z22v 1201 16.6 1.058 1.081 1.058 1.081 jit_kernel_multiply 6 15.6 0.841 1.068 0.841 1.068 make_basis_sm 11 9.8 0.002 0.005 1.061 1.062 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.980 1.044 multiply_cannon_sync_h2d 9144 15.5 1.040 1.042 1.040 1.042 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.987 1.019 acc_transpose_blocks 9144 15.5 0.037 0.037 0.908 0.954 qs_create_task_list 11 7.9 0.000 0.000 0.939 0.949 generate_qs_task_list 11 8.9 0.372 0.393 0.938 0.948 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.852 0.865 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.186000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=705.727273, yerr=14.058908 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.632448E+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 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 510505. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.027 84.425 84.426 qs_mol_dyn_low 1 2.0 0.003 0.003 84.174 84.184 qs_forces 11 3.9 0.002 0.002 84.109 84.109 qs_energies 11 4.9 0.005 0.007 81.278 81.287 scf_env_do_scf 11 5.9 0.000 0.001 72.321 72.323 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 66.848 66.848 dbcsr_multiply_generic 2055 12.4 0.107 0.111 53.025 53.232 qs_scf_new_mos 99 7.5 0.000 0.001 49.286 49.419 qs_scf_loop_do_ot 99 8.5 0.001 0.001 49.286 49.418 ot_scf_mini 99 9.5 0.002 0.002 46.711 46.782 multiply_cannon 2055 13.4 0.177 0.187 42.976 43.739 multiply_cannon_loop 2055 14.4 1.564 1.601 41.883 42.690 velocity_verlet 10 3.0 0.001 0.001 42.505 42.505 ot_mini 99 10.5 0.001 0.001 27.750 27.855 qs_ot_get_derivative 99 11.5 0.001 0.001 20.843 20.958 multiply_cannon_multrec 49320 15.4 12.078 12.818 17.104 17.703 rebuild_ks_matrix 110 8.3 0.000 0.001 14.532 14.646 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.532 14.646 mp_waitall_1 241148 16.1 12.510 13.464 12.510 13.464 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.812 12.904 multiply_cannon_sync_h2d 49320 15.4 9.888 10.409 9.888 10.409 qs_ot_get_p 110 10.4 0.001 0.001 10.039 10.169 multiply_cannon_metrocomm3 49320 15.4 0.080 0.083 6.979 8.247 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.210 7.679 apply_single 110 13.6 0.000 0.001 7.210 7.678 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.018 7.549 sum_up_and_integrate 110 10.3 0.037 0.044 7.078 7.093 integrate_v_rspace 110 11.3 0.002 0.003 7.041 7.064 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.706 6.838 calculate_rho_elec 110 8.6 0.020 0.024 6.705 6.838 init_scf_run 11 5.9 0.000 0.001 6.807 6.807 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.807 6.807 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.622 6.690 ot_diis_step 99 11.5 0.005 0.006 6.623 6.623 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.283 6.335 mp_sum_l 6514 12.8 5.301 6.129 5.301 6.129 init_scf_loop 11 6.9 0.000 0.000 5.418 5.418 dbcsr_mm_accdrv_process 87628 16.1 2.093 2.182 4.905 5.195 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.185 5.191 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.646 4.647 cp_fm_redistribute_end 48 14.0 4.041 4.620 4.045 4.622 cp_fm_diag_elpa_base 48 14.0 0.567 4.545 0.572 4.576 make_m2s 4110 13.4 0.061 0.065 4.284 4.435 make_images 4110 14.4 0.177 0.191 4.188 4.343 multiply_cannon_metrocomm1 49320 15.4 0.058 0.061 3.082 4.265 rs_pw_transfer 902 11.9 0.012 0.013 3.998 4.194 calculate_dm_sparse 110 9.5 0.000 0.001 3.935 4.058 density_rs2pw 110 9.6 0.004 0.004 3.774 3.982 wfi_extrapolate 11 7.9 0.001 0.001 3.917 3.917 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.598 3.665 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.641 3.644 grid_integrate_task_list 110 12.3 3.257 3.389 3.257 3.389 pw_transfer 1331 11.6 0.053 0.062 3.208 3.275 prepare_preconditioner 11 7.9 0.000 0.000 3.214 3.235 make_preconditioner 11 8.9 0.000 0.000 3.214 3.235 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.170 3.213 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.121 3.186 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.005 3.058 calculate_first_density_matrix 1 7.0 0.000 0.000 2.807 2.811 fft3d_ps 1111 14.6 0.747 0.833 2.657 2.719 potential_pw2rs 110 12.3 0.006 0.007 2.669 2.690 fft_wrap_pw1pw2_140 451 13.1 0.170 0.192 2.567 2.634 jit_kernel_multiply 13 15.9 2.529 2.540 2.529 2.540 mp_alltoall_d11v 2046 13.8 2.132 2.515 2.132 2.515 acc_transpose_blocks 49320 15.4 0.221 0.231 2.189 2.255 mp_waitany 14300 13.8 1.908 2.226 1.908 2.226 grid_collocate_task_list 110 9.6 2.085 2.197 2.085 2.197 make_images_data 4110 15.4 0.043 0.047 1.967 2.166 mp_sum_d 3879 11.9 1.571 2.071 1.571 2.071 hybrid_alltoall_any 4261 16.3 0.083 0.483 1.686 2.025 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.865 1.885 cp_fm_cholesky_invert 11 10.9 1.780 1.783 1.780 1.783 mp_alltoall_z22v 1111 16.6 1.589 1.737 1.589 1.737 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.426000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.727273, yerr=3.332782 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 587.902976E+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 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 1935297. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 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 0.030 0.093 69.977 69.978 qs_mol_dyn_low 1 2.0 0.003 0.003 69.634 69.643 qs_forces 11 3.9 0.002 0.002 69.537 69.538 qs_energies 11 4.9 0.002 0.006 66.256 66.259 scf_env_do_scf 11 5.9 0.001 0.001 57.611 57.615 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 49.270 49.272 dbcsr_multiply_generic 2055 12.4 0.113 0.117 37.996 38.160 velocity_verlet 10 3.0 0.001 0.002 37.041 37.043 qs_scf_new_mos 99 7.5 0.001 0.001 32.892 33.037 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.891 33.037 multiply_cannon 2055 13.4 0.223 0.244 31.245 32.518 ot_scf_mini 99 9.5 0.003 0.003 31.198 31.340 multiply_cannon_loop 2055 14.4 0.922 0.945 29.937 30.993 ot_mini 99 10.5 0.001 0.001 18.281 18.432 multiply_cannon_multrec 24660 15.4 7.624 9.347 13.913 15.482 rebuild_ks_matrix 110 8.3 0.000 0.001 13.751 13.868 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.750 13.867 qs_ot_get_derivative 99 11.5 0.001 0.001 12.476 12.620 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.151 12.251 mp_waitall_1 186928 16.3 8.247 11.111 8.247 11.111 init_scf_loop 11 6.9 0.001 0.002 8.306 8.310 multiply_cannon_sync_h2d 24660 15.4 7.028 8.178 7.028 8.178 multiply_cannon_metrocomm3 24660 15.4 0.069 0.071 5.240 7.960 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.514 7.280 apply_single 110 13.6 0.000 0.001 6.514 7.280 sum_up_and_integrate 110 10.3 0.052 0.059 6.607 6.621 integrate_v_rspace 110 11.3 0.002 0.002 6.555 6.568 dbcsr_mm_accdrv_process 52282 16.1 4.644 5.450 6.129 6.376 qs_ot_get_p 110 10.4 0.001 0.001 6.155 6.310 prepare_preconditioner 11 7.9 0.000 0.000 6.262 6.281 make_preconditioner 11 8.9 0.000 0.001 6.262 6.281 init_scf_run 11 5.9 0.000 0.002 6.273 6.274 scf_env_initial_rho_setup 11 6.9 0.002 0.007 6.272 6.273 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.190 6.203 calculate_rho_elec 110 8.6 0.039 0.047 6.190 6.202 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.852 6.011 ot_diis_step 99 11.5 0.010 0.010 5.759 5.759 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.778 5.509 make_m2s 4110 13.4 0.057 0.060 4.306 4.793 make_images 4110 14.4 0.398 0.441 4.201 4.686 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.178 4.197 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.750 3.751 density_rs2pw 110 9.6 0.004 0.004 3.442 3.618 pw_transfer 1331 11.6 0.066 0.069 3.469 3.610 wfi_extrapolate 11 7.9 0.001 0.001 3.524 3.524 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.362 3.500 rs_pw_transfer 902 11.9 0.012 0.014 3.259 3.445 cp_fm_cholesky_invert 11 10.9 3.371 3.377 3.371 3.377 grid_integrate_task_list 110 12.3 3.143 3.321 3.143 3.321 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.287 3.289 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.118 3.202 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.173 3.174 cp_fm_redistribute_end 48 14.0 2.364 3.149 2.366 3.149 cp_fm_diag_elpa_base 48 14.0 0.748 3.033 0.780 3.120 calculate_dm_sparse 110 9.5 0.002 0.012 3.019 3.057 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.938 2.986 fft_wrap_pw1pw2_140 451 13.1 0.201 0.220 2.818 2.956 make_images_data 4110 15.4 0.047 0.051 2.428 2.943 hybrid_alltoall_any 4261 16.3 0.103 0.445 2.127 2.904 fft3d_ps 1111 14.6 1.073 1.270 2.719 2.844 calculate_first_density_matrix 1 7.0 0.003 0.018 2.661 2.666 potential_pw2rs 110 12.3 0.008 0.009 2.458 2.476 mp_sum_l 6514 12.8 1.863 2.428 1.863 2.428 grid_collocate_task_list 110 9.6 2.119 2.296 2.119 2.296 mp_alltoall_d11v 2046 13.8 1.830 2.108 1.830 2.108 jit_kernel_multiply 10 16.4 1.138 2.053 1.138 2.053 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.021 2.041 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.859 1.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.763 1.777 mp_allgather_i34 2055 14.4 0.637 1.678 0.637 1.678 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.780 1.665 acc_transpose_blocks 24660 15.4 0.109 0.112 1.536 1.577 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.555 1.570 mp_irecv_dv 57340 16.2 0.656 1.550 0.656 1.550 mp_waitany 10164 13.8 1.315 1.508 1.315 1.508 mp_alltoall_z22v 1111 16.6 1.392 1.474 1.392 1.474 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.367 1.472 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.978000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.818182, yerr=6.846607 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 660.135936E+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 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 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 0.038 0.046 61.190 61.192 qs_mol_dyn_low 1 2.0 0.003 0.004 60.776 60.784 qs_forces 11 3.9 0.002 0.003 60.677 60.677 qs_energies 11 4.9 0.050 0.051 57.564 57.567 scf_env_do_scf 11 5.9 0.001 0.001 49.532 49.533 scf_env_do_scf_inner_loop 99 6.5 0.006 0.017 39.537 39.537 velocity_verlet 10 3.0 0.001 0.001 32.602 32.604 dbcsr_multiply_generic 2055 12.4 0.107 0.110 28.236 28.479 qs_scf_new_mos 99 7.5 0.001 0.001 24.848 24.936 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.847 24.935 ot_scf_mini 99 9.5 0.003 0.003 23.603 23.721 multiply_cannon 2055 13.4 0.212 0.225 22.010 23.171 multiply_cannon_loop 2055 14.4 0.614 0.632 20.817 21.991 ot_mini 99 10.5 0.001 0.001 13.290 13.400 rebuild_ks_matrix 110 8.3 0.000 0.000 12.131 12.256 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.131 12.256 multiply_cannon_multrec 16440 15.4 3.950 5.334 9.684 11.014 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.715 10.826 mp_waitall_1 146766 16.3 7.144 10.378 7.144 10.378 init_scf_loop 11 6.9 0.001 0.002 9.963 9.965 qs_ot_get_derivative 99 11.5 0.001 0.001 8.894 9.006 prepare_preconditioner 11 7.9 0.000 0.000 8.228 8.241 make_preconditioner 11 8.9 0.000 0.001 8.228 8.241 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.573 7.913 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.216 7.267 sum_up_and_integrate 110 10.3 0.059 0.061 6.375 6.392 integrate_v_rspace 110 11.3 0.002 0.003 6.315 6.332 dbcsr_mm_accdrv_process 34862 16.1 4.607 5.360 5.589 5.753 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.712 5.721 calculate_rho_elec 110 8.6 0.058 0.058 5.711 5.721 qs_ot_get_p 110 10.4 0.001 0.001 5.478 5.619 init_scf_run 11 5.9 0.000 0.001 5.550 5.551 scf_env_initial_rho_setup 11 6.9 0.002 0.005 5.550 5.551 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.958 5.378 apply_single 110 13.6 0.000 0.000 4.958 5.378 make_m2s 4110 13.4 0.051 0.053 4.143 4.526 make_images 4110 14.4 0.392 0.508 4.031 4.414 ot_diis_step 99 11.5 0.011 0.011 4.370 4.370 multiply_cannon_sync_h2d 16440 15.4 3.690 4.277 3.690 4.277 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.716 3.721 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.049 3.706 cp_fm_cholesky_invert 11 10.9 3.696 3.702 3.696 3.702 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.377 3.377 grid_integrate_task_list 110 12.3 3.190 3.364 3.190 3.364 pw_transfer 1331 11.6 0.065 0.071 3.233 3.239 density_rs2pw 110 9.6 0.004 0.004 2.966 3.177 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.126 3.135 rs_pw_transfer 902 11.9 0.011 0.012 2.768 2.990 wfi_extrapolate 11 7.9 0.001 0.001 2.947 2.947 make_images_data 4110 15.4 0.044 0.047 2.417 2.909 hybrid_alltoall_any 4261 16.3 0.105 0.378 2.109 2.824 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.808 2.814 cp_fm_redistribute_end 48 14.0 1.746 2.791 1.750 2.793 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.762 2.763 cp_fm_diag_elpa_base 48 14.0 0.976 2.651 1.033 2.759 fft_wrap_pw1pw2_140 451 13.1 0.210 0.215 2.677 2.687 calculate_dm_sparse 110 9.5 0.001 0.001 2.519 2.552 calculate_first_density_matrix 1 7.0 0.008 0.060 2.521 2.523 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.425 2.468 fft3d_ps 1111 14.6 1.066 1.078 2.441 2.451 multiply_cannon_metrocomm4 14385 15.4 0.044 0.048 0.874 2.359 grid_collocate_task_list 110 9.6 2.181 2.345 2.181 2.345 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.257 2.315 mp_irecv_dv 48980 15.7 0.805 2.238 0.805 2.238 potential_pw2rs 110 12.3 0.010 0.010 2.204 2.214 cp_fm_cholesky_decompose 22 10.9 2.070 2.084 2.070 2.084 mp_sum_l 6514 12.8 1.478 2.050 1.478 2.050 mp_alltoall_d11v 2046 13.8 1.744 2.041 1.744 2.041 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.009 2.010 dbcsr_complete_redistribute 325 12.2 0.337 0.372 1.425 1.861 cp_fm_upper_to_full 70 13.6 1.370 1.833 1.370 1.833 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.617 1.631 mp_allgather_i34 2055 14.4 0.519 1.554 0.519 1.554 jit_kernel_multiply 8 16.5 0.599 1.468 0.599 1.468 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.453 1.466 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.353 1.441 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.958 1.401 mp_waitany 17072 13.8 1.165 1.362 1.165 1.362 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.279 1.294 acc_transpose_blocks 16440 15.4 0.071 0.072 1.215 1.229 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.192000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.545455, yerr=8.648317 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 729.485312E+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 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.028 65.785 65.785 qs_mol_dyn_low 1 2.0 0.003 0.003 65.581 65.590 qs_forces 11 3.9 0.002 0.002 65.516 65.517 qs_energies 11 4.9 0.001 0.002 62.168 62.172 scf_env_do_scf 11 5.9 0.000 0.001 53.233 53.236 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.601 41.601 velocity_verlet 10 3.0 0.001 0.001 37.488 37.490 dbcsr_multiply_generic 2055 12.4 0.114 0.116 30.385 30.600 qs_scf_new_mos 99 7.5 0.001 0.001 26.797 26.916 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.796 26.915 ot_scf_mini 99 9.5 0.003 0.003 25.147 25.257 multiply_cannon 2055 13.4 0.247 0.266 23.031 23.936 multiply_cannon_loop 2055 14.4 0.884 0.899 21.640 22.168 ot_mini 99 10.5 0.001 0.001 14.833 14.967 multiply_cannon_multrec 24660 15.4 4.248 6.797 12.762 14.073 rebuild_ks_matrix 110 8.3 0.000 0.000 12.293 12.383 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.292 12.383 init_scf_loop 11 6.9 0.000 0.000 11.593 11.593 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.927 11.009 qs_ot_get_derivative 99 11.5 0.001 0.001 10.694 10.812 prepare_preconditioner 11 7.9 0.000 0.000 9.783 9.804 make_preconditioner 11 8.9 0.000 0.000 9.783 9.804 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.971 9.456 dbcsr_mm_accdrv_process 52304 16.0 6.904 7.950 8.368 9.342 mp_waitall_1 126806 16.4 5.222 7.079 5.222 7.079 sum_up_and_integrate 110 10.3 0.066 0.070 6.598 6.613 integrate_v_rspace 110 11.3 0.002 0.003 6.532 6.543 init_scf_run 11 5.9 0.000 0.001 6.238 6.238 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.238 6.238 make_m2s 4110 13.4 0.060 0.061 5.681 5.965 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.820 5.828 calculate_rho_elec 110 8.6 0.077 0.080 5.820 5.828 make_images 4110 14.4 0.577 0.694 5.542 5.823 qs_ot_get_p 110 10.4 0.001 0.001 5.333 5.486 cp_fm_upper_to_full 70 13.8 3.325 4.730 3.325 4.730 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.036 4.141 apply_single 110 13.6 0.000 0.000 4.035 4.141 ot_diis_step 99 11.5 0.011 0.011 4.104 4.104 dbcsr_complete_redistribute 325 12.2 0.422 0.462 2.864 4.025 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.551 3.617 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.602 3.616 pw_transfer 1331 11.6 0.064 0.071 3.505 3.530 grid_integrate_task_list 110 12.3 3.283 3.439 3.283 3.439 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.400 3.427 multiply_cannon_sync_h2d 24660 15.4 3.206 3.391 3.206 3.391 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.243 3.387 wfi_extrapolate 11 7.9 0.001 0.001 3.386 3.387 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.574 3.377 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.262 3.320 make_images_data 4110 15.4 0.046 0.050 2.943 3.288 hybrid_alltoall_any 4261 16.3 0.119 0.454 2.473 3.249 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.240 3.242 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.130 3.130 density_rs2pw 110 9.6 0.004 0.004 2.889 3.051 calculate_dm_sparse 110 9.5 0.001 0.001 2.955 2.991 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.803 2.924 mp_alltoall_i22 605 13.7 1.696 2.892 1.696 2.892 fft_wrap_pw1pw2_140 451 13.1 0.202 0.213 2.827 2.859 calculate_first_density_matrix 1 7.0 0.000 0.000 2.736 2.740 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.701 2.729 rs_pw_transfer 902 11.9 0.010 0.011 2.534 2.726 fft3d_ps 1111 14.6 1.063 1.098 2.703 2.723 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.616 2.618 cp_fm_redistribute_end 48 14.0 1.301 2.592 1.303 2.593 cp_fm_diag_elpa_base 48 14.0 1.208 2.464 1.287 2.576 cp_fm_cholesky_invert 11 10.9 2.567 2.575 2.567 2.575 mp_alltoall_d11v 2046 13.8 2.114 2.455 2.114 2.455 grid_collocate_task_list 110 9.6 2.222 2.356 2.222 2.356 potential_pw2rs 110 12.3 0.012 0.013 2.313 2.331 jit_kernel_multiply 12 15.9 1.135 2.248 1.135 2.248 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.217 2.218 cp_fm_cholesky_decompose 22 10.9 1.702 1.762 1.702 1.762 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.672 1.709 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.598 1.696 mp_sum_l 6514 12.8 0.985 1.665 0.985 1.665 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.614 1.625 acc_transpose_blocks 24660 15.4 0.102 0.104 1.536 1.583 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.540 1.556 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.827 1.545 mp_alltoall_z22v 1111 16.6 1.424 1.492 1.424 1.492 mp_irecv_dv 62702 16.1 0.729 1.468 0.729 1.468 mp_allgather_i34 2055 14.4 0.488 1.460 0.488 1.460 mp_waitany 13376 13.8 1.123 1.341 1.123 1.341 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.785000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=688.363636, yerr=8.423697 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 827.162624E+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 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 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 0.032 56.723 56.724 qs_mol_dyn_low 1 2.0 0.003 0.003 56.498 56.507 qs_forces 11 3.9 0.002 0.002 56.398 56.408 qs_energies 11 4.9 0.004 0.005 52.810 52.824 scf_env_do_scf 11 5.9 0.000 0.001 43.974 43.975 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 36.210 36.211 velocity_verlet 10 3.0 0.001 0.001 31.201 31.206 dbcsr_multiply_generic 2055 12.4 0.104 0.106 23.421 23.525 qs_scf_new_mos 99 7.5 0.001 0.001 21.892 21.940 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.892 21.940 ot_scf_mini 99 9.5 0.002 0.002 20.596 20.621 multiply_cannon 2055 13.4 0.247 0.259 17.662 19.076 multiply_cannon_loop 2055 14.4 0.321 0.332 16.222 16.529 rebuild_ks_matrix 110 8.3 0.000 0.000 11.553 11.582 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.553 11.581 ot_mini 99 10.5 0.001 0.001 10.683 10.711 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.322 10.350 multiply_cannon_multrec 8220 15.4 3.251 4.581 7.544 8.766 mp_waitall_1 106626 16.5 6.663 8.509 6.663 8.509 init_scf_loop 11 6.9 0.000 0.000 7.714 7.716 qs_ot_get_derivative 99 11.5 0.001 0.001 6.840 6.861 sum_up_and_integrate 110 10.3 0.079 0.081 6.223 6.236 integrate_v_rspace 110 11.3 0.003 0.003 6.143 6.155 prepare_preconditioner 11 7.9 0.000 0.000 6.081 6.085 make_preconditioner 11 8.9 0.000 0.000 6.081 6.085 qs_ot_get_p 110 10.4 0.001 0.001 5.867 5.885 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.730 5.744 calculate_rho_elec 110 8.6 0.115 0.115 5.729 5.743 init_scf_run 11 5.9 0.000 0.001 5.733 5.733 scf_env_initial_rho_setup 11 6.9 0.002 0.002 5.733 5.733 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.657 5.733 dbcsr_mm_accdrv_process 17442 15.9 2.818 3.803 4.162 5.044 make_m2s 4110 13.4 0.040 0.041 4.439 4.795 make_images 4110 14.4 0.637 0.690 4.309 4.664 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.475 4.480 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.007 4.294 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.132 4.133 ot_diis_step 99 11.5 0.012 0.012 3.818 3.818 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.724 3.741 apply_single 110 13.6 0.000 0.000 3.724 3.740 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.542 3.543 cp_fm_redistribute_end 48 14.0 0.893 3.507 0.904 3.508 grid_integrate_task_list 110 12.3 3.360 3.489 3.360 3.489 cp_fm_diag_elpa_base 48 14.0 2.416 3.296 2.595 3.464 pw_transfer 1331 11.6 0.064 0.069 3.308 3.320 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.201 3.216 make_images_data 4110 15.4 0.039 0.043 2.570 3.108 multiply_cannon_sync_h2d 8220 15.4 2.920 3.044 2.920 3.044 cp_fm_cholesky_invert 11 10.9 2.957 2.961 2.957 2.961 hybrid_alltoall_any 4261 16.3 0.200 0.863 2.463 2.935 calculate_first_density_matrix 1 7.0 0.000 0.000 2.925 2.926 density_rs2pw 110 9.6 0.004 0.004 2.770 2.893 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.743 2.744 fft_wrap_pw1pw2_140 451 13.1 0.212 0.215 2.711 2.727 wfi_extrapolate 11 7.9 0.001 0.001 2.719 2.720 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.645 2.655 calculate_dm_sparse 110 9.5 0.001 0.001 2.529 2.568 fft3d_ps 1111 14.6 1.113 1.143 2.460 2.470 rs_pw_transfer 902 11.9 0.010 0.011 2.311 2.461 grid_collocate_task_list 110 9.6 2.327 2.432 2.327 2.432 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.267 2.282 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.127 2.151 potential_pw2rs 110 12.3 0.015 0.016 1.984 1.989 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.769 1.988 mp_alltoall_d11v 2046 13.8 1.720 1.890 1.720 1.890 mp_allgather_i34 2055 14.4 0.627 1.825 0.627 1.825 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.802 1.812 cp_fm_cholesky_decompose 22 10.9 1.741 1.772 1.741 1.772 jit_kernel_multiply 9 15.7 1.042 1.633 1.042 1.633 dbcsr_complete_redistribute 325 12.2 0.562 0.594 1.527 1.616 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.490 1.612 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.567 1.572 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.434 1.449 multiply_cannon_metrocomm1 8220 15.4 0.021 0.021 0.983 1.331 qs_create_task_list 11 7.9 0.000 0.001 1.223 1.317 generate_qs_task_list 11 8.9 0.379 0.445 1.222 1.317 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.230 1.254 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.477 1.246 make_basis_sm 11 9.8 0.001 0.003 1.206 1.211 mp_irecv_dv 24056 15.7 0.453 1.206 0.453 1.206 mp_waitany 9240 13.8 1.044 1.196 1.044 1.196 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.724000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=782.545455, yerr=10.439119 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 1.391436E+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_ISendRecv 4620 360267. MP_Wait 7524 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 0.019 0.053 89.954 89.955 qs_mol_dyn_low 1 2.0 0.003 0.003 89.556 89.564 qs_forces 11 3.9 0.002 0.002 89.481 89.493 qs_energies 11 4.9 0.035 0.039 85.453 85.467 scf_env_do_scf 11 5.9 0.000 0.001 75.543 75.543 velocity_verlet 10 3.0 0.001 0.001 57.292 57.298 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.378 45.379 init_scf_loop 11 6.9 0.000 0.000 30.088 30.090 dbcsr_multiply_generic 2055 12.4 0.124 0.127 29.274 29.379 qs_scf_new_mos 99 7.5 0.001 0.001 28.130 28.177 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.129 28.176 prepare_preconditioner 11 7.9 0.000 0.000 28.150 28.157 make_preconditioner 11 8.9 0.000 0.000 28.150 28.157 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.349 27.633 ot_scf_mini 99 9.5 0.002 0.002 26.290 26.328 multiply_cannon 2055 13.4 0.352 0.376 21.741 22.563 multiply_cannon_loop 2055 14.4 0.339 0.342 19.809 20.122 cp_fm_upper_to_full 70 14.2 12.777 18.304 12.777 18.304 ot_mini 99 10.5 0.001 0.001 14.182 14.219 rebuild_ks_matrix 110 8.3 0.000 0.001 13.711 13.751 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.710 13.751 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.536 12.572 dbcsr_complete_redistribute 325 12.2 1.025 1.047 7.499 10.759 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.373 9.634 qs_ot_get_derivative 99 11.5 0.001 0.001 9.580 9.621 multiply_cannon_multrec 8220 15.4 4.373 4.565 9.519 9.605 mp_waitall_1 87304 16.6 8.439 9.605 8.439 9.605 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.785 9.023 mp_alltoall_i22 605 13.7 5.451 8.701 5.451 8.701 sum_up_and_integrate 110 10.3 0.151 0.152 7.056 7.070 integrate_v_rspace 110 11.3 0.004 0.004 6.905 6.919 qs_ot_get_p 110 10.4 0.001 0.001 6.809 6.854 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.734 6.764 calculate_rho_elec 110 8.6 0.227 0.227 6.733 6.763 cp_fm_cholesky_invert 11 10.9 6.224 6.228 6.224 6.228 make_m2s 4110 13.4 0.044 0.044 5.609 6.205 make_images 4110 14.4 0.878 0.927 5.420 6.017 init_scf_run 11 5.9 0.000 0.001 5.595 5.595 scf_env_initial_rho_setup 11 6.9 0.014 0.015 5.595 5.595 dbcsr_mm_accdrv_process 11614 15.7 3.246 3.649 5.004 5.258 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.888 5.200 qs_ot_p2m_diag 48 11.0 0.151 0.156 5.149 5.165 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.687 5.163 apply_single 110 13.6 0.000 0.000 4.687 5.163 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.641 4.642 ot_diis_step 99 11.5 0.015 0.015 4.578 4.578 pw_transfer 1331 11.6 0.073 0.073 4.130 4.138 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.015 4.023 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.965 3.965 cp_fm_diag_elpa_base 48 14.0 3.421 3.614 3.963 3.963 multiply_cannon_sync_h2d 8220 15.4 3.948 3.953 3.948 3.953 hybrid_alltoall_any 4261 16.3 0.255 0.550 3.051 3.882 make_images_data 4110 15.4 0.042 0.044 3.110 3.851 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.342 3.772 grid_integrate_task_list 110 12.3 3.661 3.711 3.661 3.711 qs_energies_init_hamiltonians 11 5.9 0.017 0.018 3.654 3.666 fft_wrap_pw1pw2_140 451 13.1 0.214 0.217 3.456 3.462 fft3d_ps 1111 14.6 1.262 1.271 3.224 3.229 wfi_extrapolate 11 7.9 0.001 0.001 3.227 3.227 calculate_dm_sparse 110 9.5 0.001 0.001 3.191 3.216 density_rs2pw 110 9.6 0.004 0.004 3.183 3.203 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.868 2.871 cp_fm_cholesky_decompose 22 10.9 2.745 2.761 2.745 2.761 grid_collocate_task_list 110 9.6 2.630 2.658 2.630 2.658 rs_pw_transfer 902 11.9 0.010 0.010 2.395 2.444 potential_pw2rs 110 12.3 0.021 0.022 2.419 2.430 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.382 2.410 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.352 2.364 mp_alltoall_d11v 2046 13.8 2.302 2.348 2.302 2.348 calculate_first_density_matrix 1 7.0 0.000 0.000 2.251 2.252 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.170 2.232 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.081 2.181 qs_create_task_list 11 7.9 0.000 0.000 1.896 1.939 generate_qs_task_list 11 8.9 0.737 0.791 1.896 1.939 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.885 1.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.955000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1232.909091, yerr=73.460508 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 627.634176E+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 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 2980309. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 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 0.015 0.037 210.460 210.461 qs_mol_dyn_low 1 2.0 0.003 0.003 210.101 210.114 qs_forces 11 3.9 0.003 0.003 209.980 209.980 qs_energies 11 4.9 0.001 0.002 204.377 204.399 scf_env_do_scf 11 5.9 0.001 0.001 187.450 187.453 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 165.621 165.630 velocity_verlet 10 3.0 0.001 0.001 126.237 126.238 dbcsr_multiply_generic 2507 12.6 0.175 0.180 125.764 126.216 qs_scf_new_mos 117 7.6 0.001 0.001 124.373 124.607 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.372 124.606 ot_scf_mini 117 9.6 0.003 0.003 117.726 117.965 multiply_cannon 2507 13.6 0.238 0.246 101.464 103.060 multiply_cannon_loop 2507 14.6 2.130 2.204 98.998 100.607 ot_mini 117 10.6 0.001 0.001 66.807 67.033 multiply_cannon_multrec 60168 15.6 32.828 34.884 41.219 43.067 qs_ot_get_derivative 117 11.6 0.001 0.001 42.160 42.360 rebuild_ks_matrix 128 8.3 0.001 0.001 34.450 34.640 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 34.450 34.639 mp_waitall_1 291448 16.2 30.969 33.740 30.969 33.740 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.015 31.160 qs_ot_get_p 128 10.4 0.001 0.001 28.967 29.239 multiply_cannon_sync_h2d 60168 15.6 27.155 28.541 27.155 28.541 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.129 24.868 apply_single 128 13.6 0.001 0.001 24.129 24.868 ot_diis_step 117 11.6 0.007 0.008 24.335 24.336 qs_ot_p2m_diag 83 11.4 0.077 0.091 22.057 22.166 init_scf_loop 11 6.9 0.000 0.001 21.644 21.645 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.912 20.082 cp_dbcsr_syevd 83 12.4 0.004 0.005 19.362 19.365 multiply_cannon_metrocomm3 60168 15.6 0.117 0.121 15.663 17.618 prepare_preconditioner 11 7.9 0.000 0.000 17.091 17.138 make_preconditioner 11 8.9 0.000 0.000 17.091 17.138 make_full_inverse_cholesky 11 9.9 0.000 0.000 16.320 16.484 cp_fm_diag_elpa 83 13.4 0.000 0.001 16.385 16.391 cp_fm_redistribute_end 83 14.4 12.891 16.280 12.905 16.285 cp_fm_diag_elpa_base 83 14.4 3.326 16.034 3.364 16.167 make_m2s 5014 13.6 0.103 0.112 14.684 15.053 sum_up_and_integrate 128 10.3 0.090 0.107 14.935 14.950 make_images 5014 14.6 0.405 0.422 14.509 14.889 integrate_v_rspace 128 11.3 0.003 0.004 14.845 14.861 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.057 14.150 calculate_rho_elec 128 8.7 0.045 0.063 14.056 14.149 init_scf_run 11 5.9 0.000 0.001 12.505 12.506 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.505 12.505 cp_fm_cholesky_invert 11 10.9 9.765 9.773 9.765 9.773 mp_sum_l 7870 13.0 8.352 9.464 8.352 9.464 wfi_extrapolate 11 7.9 0.001 0.001 9.010 9.010 calculate_dm_sparse 128 9.5 0.001 0.001 8.502 8.618 multiply_cannon_metrocomm1 60168 15.6 0.086 0.089 6.851 8.601 dbcsr_mm_accdrv_process 124484 16.2 3.283 3.466 7.950 8.494 density_rs2pw 128 9.7 0.006 0.007 7.790 8.294 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.144 8.276 make_images_data 5014 15.6 0.071 0.077 7.209 8.250 pw_transfer 1547 11.6 0.075 0.107 7.755 8.041 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.903 8.022 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 7.551 7.810 hybrid_alltoall_any 5200 16.5 0.294 2.276 6.273 7.654 grid_integrate_task_list 128 12.3 7.142 7.516 7.142 7.516 rs_pw_transfer 1046 11.9 0.017 0.019 6.968 7.469 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.847 6.856 fft3d_ps 1291 14.7 2.088 2.570 6.342 6.543 fft_wrap_pw1pw2_140 523 13.2 0.445 0.512 6.131 6.342 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.904 5.989 mp_alltoall_d11v 2415 14.1 4.534 5.735 4.534 5.735 potential_pw2rs 128 12.3 0.009 0.010 5.288 5.346 cp_fm_cholesky_decompose 22 10.9 5.136 5.150 5.136 5.150 grid_collocate_task_list 128 9.7 4.728 5.108 4.728 5.108 mp_sum_d 4460 12.2 3.564 4.421 3.564 4.421 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=210.461000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.909091, yerr=6.556808 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/16/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 2.166473E+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 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 826.941440E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069299E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842592400 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 57283. MP_Allreduce 11237 947. MP_Sync 168 MP_Alltoall 1955 5704603. MP_ISendRecv 23876 47072. MP_Wait 37656 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.031 192.585 192.586 qs_mol_dyn_low 1 2.0 0.003 0.003 191.959 191.971 qs_forces 11 3.9 0.037 0.043 191.862 191.863 qs_energies 11 4.9 0.001 0.002 185.053 185.096 scf_env_do_scf 11 5.9 0.001 0.001 167.975 167.985 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 134.442 134.444 velocity_verlet 10 3.0 0.001 0.001 120.541 120.543 dbcsr_multiply_generic 2485 12.5 0.184 0.189 98.292 99.460 qs_scf_new_mos 116 7.6 0.001 0.001 95.584 96.068 qs_scf_loop_do_ot 116 8.6 0.001 0.001 95.583 96.067 ot_scf_mini 116 9.6 0.004 0.004 90.810 91.384 multiply_cannon 2485 13.5 0.476 0.528 77.601 82.156 multiply_cannon_loop 2485 14.5 1.237 1.274 74.182 76.455 ot_mini 116 10.6 0.001 0.001 50.529 51.010 mp_waitall_1 224796 16.4 25.767 39.850 25.767 39.850 multiply_cannon_multrec 29820 15.5 21.852 26.713 31.538 36.614 init_scf_loop 11 6.9 0.000 0.000 33.299 33.300 rebuild_ks_matrix 127 8.3 0.001 0.001 32.547 33.002 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 32.547 33.002 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.304 29.723 qs_ot_get_derivative 116 11.6 0.001 0.002 28.702 29.236 prepare_preconditioner 11 7.9 0.000 0.000 28.940 29.009 make_preconditioner 11 8.9 0.000 0.000 28.940 29.009 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.677 28.226 multiply_cannon_metrocomm3 29820 15.5 0.093 0.099 15.796 28.037 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.835 23.103 apply_single 127 13.6 0.001 0.001 21.835 23.102 qs_ot_get_p 127 10.4 0.001 0.001 22.002 22.674 ot_diis_step 116 11.6 0.014 0.015 21.652 21.654 multiply_cannon_sync_h2d 29820 15.5 19.175 21.512 19.175 21.512 cp_fm_cholesky_invert 11 10.9 16.960 16.973 16.960 16.973 qs_ot_p2m_diag 82 11.4 0.185 0.213 16.645 16.678 make_m2s 4970 13.5 0.091 0.096 14.866 16.519 make_images 4970 14.5 1.144 1.335 14.665 16.311 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.438 15.439 sum_up_and_integrate 127 10.3 0.114 0.131 14.250 14.275 integrate_v_rspace 127 11.3 0.003 0.004 14.136 14.165 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.379 13.412 calculate_rho_elec 127 8.7 0.087 0.104 13.379 13.411 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.057 12.077 cp_fm_redistribute_end 82 14.4 7.012 11.982 7.033 11.991 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.552 11.896 cp_fm_diag_elpa_base 82 14.4 4.711 11.467 4.924 11.827 init_scf_run 11 5.9 0.000 0.001 11.691 11.692 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.691 11.692 make_images_data 4970 15.5 0.066 0.072 8.926 10.959 hybrid_alltoall_any 5155 16.4 0.340 1.492 7.677 10.480 multiply_cannon_metrocomm4 27335 15.5 0.096 0.111 3.728 10.138 mp_irecv_dv 68888 16.3 3.535 9.762 3.535 9.762 dbcsr_mm_accdrv_process 61726 16.2 4.500 5.294 9.151 9.747 wfi_extrapolate 11 7.9 0.001 0.001 8.468 8.468 pw_transfer 1535 11.6 0.084 0.099 7.804 7.872 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.582 7.634 density_rs2pw 127 9.7 0.006 0.006 7.049 7.579 grid_integrate_task_list 127 12.3 7.101 7.473 7.101 7.473 cp_fm_cholesky_decompose 22 10.9 7.137 7.210 7.137 7.210 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.167 6.973 calculate_dm_sparse 127 9.5 0.001 0.001 6.599 6.762 fft_wrap_pw1pw2_140 519 13.2 0.466 0.512 6.591 6.649 rs_pw_transfer 1038 11.9 0.014 0.017 5.845 6.472 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.204 6.212 fft3d_ps 1281 14.7 2.748 2.895 5.966 6.010 mp_sum_l 7804 13.0 4.055 5.751 4.055 5.751 mp_alltoall_d11v 2401 14.1 4.698 5.588 4.698 5.588 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.356 5.421 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.299 5.411 grid_collocate_task_list 127 9.7 4.868 5.209 4.868 5.209 mp_allgather_i34 2485 14.5 2.004 5.099 2.004 5.099 potential_pw2rs 127 12.3 0.014 0.016 4.761 4.785 dbcsr_complete_redistribute 393 12.7 0.773 0.874 3.484 4.325 mp_sum_d 4439 12.1 2.690 4.024 2.690 4.024 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=192.586000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.272727, yerr=1.542778 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 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 934.424576E+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 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4083 57222. MP_Allreduce 11246 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_ISendRecv 15748 75008. MP_Wait 29528 MP_comm_split 82 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 0.067 0.129 180.095 180.096 qs_mol_dyn_low 1 2.0 0.003 0.003 179.516 179.545 qs_forces 11 3.9 0.002 0.003 179.406 179.411 qs_energies 11 4.9 0.001 0.004 172.299 172.315 scf_env_do_scf 11 5.9 0.001 0.001 156.376 156.377 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 121.490 121.491 velocity_verlet 10 3.0 0.001 0.002 114.431 114.435 dbcsr_multiply_generic 2485 12.5 0.178 0.183 83.828 84.973 qs_scf_new_mos 116 7.6 0.001 0.001 83.584 83.922 qs_scf_loop_do_ot 116 8.6 0.001 0.002 83.583 83.921 ot_scf_mini 116 9.6 0.004 0.004 79.460 79.878 multiply_cannon 2485 13.5 0.497 0.517 62.906 67.299 multiply_cannon_loop 2485 14.5 0.843 0.865 59.476 62.033 ot_mini 116 10.6 0.001 0.001 44.533 44.955 mp_waitall_1 176908 16.5 27.100 36.397 27.100 36.397 init_scf_loop 11 6.9 0.000 0.002 34.786 34.787 rebuild_ks_matrix 127 8.3 0.001 0.001 31.381 31.940 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 31.380 31.939 prepare_preconditioner 11 7.9 0.000 0.000 30.792 30.845 make_preconditioner 11 8.9 0.000 0.001 30.792 30.844 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.461 29.879 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.949 28.468 multiply_cannon_multrec 19880 15.5 13.304 16.266 22.214 25.184 multiply_cannon_metrocomm3 19880 15.5 0.060 0.064 15.747 25.044 qs_ot_get_derivative 116 11.6 0.001 0.002 24.554 25.016 qs_ot_get_p 127 10.4 0.001 0.001 20.379 20.916 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.872 20.881 apply_single 127 13.6 0.001 0.001 19.872 20.881 ot_diis_step 116 11.6 0.018 0.029 19.805 19.805 make_m2s 4970 13.5 0.082 0.087 15.171 16.151 multiply_cannon_sync_h2d 19880 15.5 14.115 15.959 14.115 15.959 make_images 4970 14.5 1.161 1.254 14.943 15.921 qs_ot_p2m_diag 82 11.4 0.262 0.270 15.709 15.716 sum_up_and_integrate 127 10.3 0.131 0.143 15.151 15.174 integrate_v_rspace 127 11.3 0.003 0.004 15.020 15.047 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.727 14.729 cp_fm_cholesky_invert 11 10.9 14.230 14.239 14.230 14.239 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.541 13.562 calculate_rho_elec 127 8.7 0.131 0.145 13.540 13.561 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.639 11.640 cp_fm_redistribute_end 82 14.4 4.364 11.568 4.379 11.572 cp_fm_diag_elpa_base 82 14.4 6.762 10.998 7.168 11.484 init_scf_run 11 5.9 0.000 0.001 10.785 10.786 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.785 10.785 make_images_data 4970 15.5 0.061 0.069 9.313 10.729 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.932 10.218 hybrid_alltoall_any 5155 16.4 0.429 1.946 8.011 9.871 multiply_cannon_metrocomm4 17395 15.5 0.058 0.068 3.444 9.405 mp_irecv_dv 49801 16.2 3.327 9.163 3.327 9.163 pw_transfer 1535 11.6 0.084 0.103 8.642 8.759 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.421 8.543 dbcsr_mm_accdrv_process 41158 16.2 4.520 5.579 8.375 8.493 grid_integrate_task_list 127 12.3 7.242 7.698 7.242 7.698 wfi_extrapolate 11 7.9 0.001 0.001 7.530 7.530 cp_fm_cholesky_decompose 22 10.9 7.476 7.506 7.476 7.506 cp_fm_upper_to_full 104 14.5 5.742 7.455 5.742 7.455 density_rs2pw 127 9.7 0.005 0.006 7.031 7.352 fft_wrap_pw1pw2_140 519 13.2 0.475 0.532 7.199 7.329 fft3d_ps 1281 14.7 2.675 2.893 6.701 6.781 dbcsr_complete_redistribute 393 12.7 1.172 1.196 4.690 6.458 rs_pw_transfer 1038 11.9 0.014 0.014 5.849 6.182 calculate_dm_sparse 127 9.5 0.001 0.001 5.861 5.982 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.666 5.672 mp_sum_l 7804 13.0 4.030 5.482 4.030 5.482 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.753 5.425 potential_pw2rs 127 12.3 0.020 0.023 5.401 5.425 grid_collocate_task_list 127 9.7 5.029 5.422 5.029 5.422 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.468 5.230 mp_alltoall_d11v 2401 14.1 4.594 5.146 4.594 5.146 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.702 4.830 mp_allgather_i34 2485 14.5 1.960 4.626 1.960 4.626 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.014 4.048 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.311 4.015 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.941 3.994 mp_alltoall_i22 712 14.1 1.985 3.896 1.985 3.896 qs_energies_init_hamiltonians 11 5.9 0.001 0.012 3.773 3.776 mp_alltoall_z22v 1281 16.7 3.435 3.630 3.435 3.630 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=180.096000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=887.454545, yerr=9.576141 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 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 11613083000832 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 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.158418E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769793E+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 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921130560 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4113 56813. MP_Allreduce 11321 1065. MP_Sync 170 MP_Alltoall 1724 12509439. MP_ISendRecv 11868 75008. MP_Wait 28546 MP_comm_split 83 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 0.060 0.128 193.380 193.381 qs_mol_dyn_low 1 2.0 0.003 0.003 192.924 192.937 qs_forces 11 3.9 0.004 0.008 192.813 192.821 qs_energies 11 4.9 0.008 0.024 185.818 185.835 scf_env_do_scf 11 5.9 0.001 0.001 168.313 168.321 velocity_verlet 10 3.0 0.001 0.001 127.526 127.529 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 120.812 120.813 qs_scf_new_mos 118 7.6 0.001 0.001 82.806 83.113 qs_scf_loop_do_ot 118 8.6 0.001 0.001 82.805 83.112 dbcsr_multiply_generic 2527 12.6 0.190 0.194 81.452 82.145 ot_scf_mini 118 9.6 0.003 0.004 78.234 78.529 multiply_cannon 2527 13.6 0.561 0.596 55.731 58.354 multiply_cannon_loop 2527 14.6 1.189 1.211 51.715 53.382 init_scf_loop 11 6.9 0.001 0.002 47.363 47.364 ot_mini 118 10.6 0.001 0.001 43.787 44.068 prepare_preconditioner 11 7.9 0.000 0.000 43.230 43.263 make_preconditioner 11 8.9 0.000 0.001 43.230 43.263 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.747 41.907 multiply_cannon_multrec 30324 15.6 14.284 19.843 26.267 31.233 rebuild_ks_matrix 129 8.3 0.001 0.001 30.622 30.908 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 30.621 30.907 mp_waitall_1 154994 16.5 19.281 29.647 19.281 29.647 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.848 28.107 qs_ot_get_derivative 118 11.6 0.001 0.002 24.056 24.351 make_m2s 5054 13.6 0.097 0.103 21.102 22.286 make_images 5054 14.6 1.967 2.252 20.806 21.994 qs_ot_get_p 129 10.4 0.001 0.001 20.564 20.897 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.103 19.675 apply_single 129 13.6 0.001 0.001 19.102 19.675 ot_diis_step 118 11.6 0.018 0.019 19.598 19.600 cp_fm_upper_to_full 105 14.7 11.475 16.996 11.475 16.996 qs_ot_p2m_diag 83 11.4 0.343 0.390 16.382 16.434 cp_fm_cholesky_invert 11 10.9 16.176 16.185 16.176 16.185 multiply_cannon_metrocomm3 30324 15.6 0.048 0.050 6.415 15.184 sum_up_and_integrate 129 10.3 0.142 0.155 15.144 15.169 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.109 15.110 integrate_v_rspace 129 11.3 0.003 0.004 15.002 15.033 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.023 14.061 calculate_rho_elec 129 8.7 0.176 0.192 14.022 14.061 dbcsr_complete_redistribute 395 12.7 1.515 1.640 9.425 13.386 make_images_data 5054 15.6 0.065 0.069 11.484 13.237 multiply_cannon_sync_h2d 30324 15.6 11.832 12.876 11.832 12.876 hybrid_alltoall_any 5240 16.5 0.529 2.228 10.249 12.405 dbcsr_mm_accdrv_process 62756 16.2 7.394 8.476 11.558 12.077 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.912 11.915 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.882 11.844 cp_fm_redistribute_end 83 14.4 2.019 11.824 2.034 11.829 cp_fm_diag_elpa_base 83 14.4 9.167 11.228 9.759 11.720 init_scf_run 11 5.9 0.000 0.001 10.831 10.834 scf_env_initial_rho_setup 11 6.9 0.006 0.021 10.831 10.833 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 10.304 10.520 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.459 10.296 mp_alltoall_i22 718 14.1 5.705 9.614 5.705 9.614 pw_transfer 1559 11.6 0.085 0.101 8.641 8.747 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.414 8.525 grid_integrate_task_list 129 12.3 7.566 7.957 7.566 7.957 cp_fm_cholesky_decompose 22 10.9 7.722 7.812 7.722 7.812 density_rs2pw 129 9.7 0.005 0.006 7.108 7.603 wfi_extrapolate 11 7.9 0.001 0.001 7.514 7.514 fft_wrap_pw1pw2_140 527 13.2 0.484 0.489 7.325 7.446 multiply_cannon_metrocomm4 25270 15.6 0.075 0.087 2.738 7.379 mp_irecv_dv 76685 16.2 2.595 7.112 2.595 7.112 fft3d_ps 1301 14.7 2.851 2.947 6.632 6.726 calculate_dm_sparse 129 9.5 0.001 0.001 6.280 6.381 rs_pw_transfer 1054 12.0 0.013 0.014 5.799 6.287 mp_alltoall_d11v 2423 14.1 5.521 5.957 5.521 5.957 grid_collocate_task_list 129 9.7 5.245 5.635 5.245 5.635 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.465 5.490 qs_energies_init_hamiltonians 11 5.9 0.008 0.021 5.087 5.093 potential_pw2rs 129 12.3 0.022 0.023 5.019 5.047 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.616 4.716 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.380 4.447 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.208 4.259 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=193.381000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1086.636364, yerr=25.677272 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/19/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 5.865089E+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 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.527693E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 58095. MP_Allreduce 11022 1174. MP_Sync 87 MP_Alltoall 1712 18838222. MP_ISendRecv 7680 122880. MP_Wait 19962 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.044 175.218 175.219 qs_mol_dyn_low 1 2.0 0.003 0.003 174.840 174.853 qs_forces 11 3.9 0.003 0.003 174.730 174.737 qs_energies 11 4.9 0.001 0.002 167.238 167.249 scf_env_do_scf 11 5.9 0.001 0.001 148.919 148.935 velocity_verlet 10 3.0 0.001 0.001 114.148 114.152 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 113.268 113.270 qs_scf_new_mos 117 7.6 0.001 0.001 76.833 76.941 qs_scf_loop_do_ot 117 8.6 0.001 0.001 76.832 76.940 dbcsr_multiply_generic 2507 12.6 0.182 0.188 74.577 75.176 ot_scf_mini 117 9.6 0.004 0.004 72.351 72.434 multiply_cannon 2507 13.6 0.586 0.617 54.575 58.604 multiply_cannon_loop 2507 14.6 0.445 0.458 49.782 50.755 ot_mini 117 10.6 0.001 0.001 39.503 39.578 init_scf_loop 11 6.9 0.000 0.000 35.484 35.486 mp_waitall_1 129618 16.6 26.496 33.480 26.496 33.480 prepare_preconditioner 11 7.9 0.000 0.000 31.672 31.709 make_preconditioner 11 8.9 0.000 0.000 31.672 31.709 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.555 29.806 rebuild_ks_matrix 128 8.3 0.001 0.001 29.013 29.101 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.012 29.101 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.341 26.418 multiply_cannon_multrec 10028 15.6 10.415 14.443 17.725 20.475 ot_diis_step 117 11.6 0.020 0.021 19.928 19.929 qs_ot_get_p 128 10.4 0.001 0.001 19.830 19.928 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.388 19.811 apply_single 128 13.6 0.001 0.001 19.387 19.811 qs_ot_get_derivative 117 11.6 0.002 0.002 19.509 19.583 multiply_cannon_metrocomm3 10028 15.6 0.023 0.024 12.016 19.300 make_m2s 5014 13.6 0.068 0.074 16.381 18.457 make_images 5014 14.6 2.322 2.861 16.075 18.155 cp_fm_cholesky_invert 11 10.9 18.077 18.083 18.077 18.083 qs_ot_p2m_diag 83 11.4 0.496 0.502 15.941 15.960 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.705 14.712 sum_up_and_integrate 128 10.3 0.181 0.191 14.368 14.412 integrate_v_rspace 128 11.3 0.004 0.004 14.187 14.241 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.684 13.719 calculate_rho_elec 128 8.7 0.258 0.269 13.683 13.718 make_images_data 5014 15.6 0.054 0.061 9.952 12.409 multiply_cannon_sync_h2d 10028 15.6 11.644 12.140 11.644 12.140 hybrid_alltoall_any 5200 16.5 0.796 3.668 9.657 12.128 init_scf_run 11 5.9 0.000 0.001 11.448 11.448 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.448 11.448 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.443 11.445 cp_fm_diag_elpa_base 83 14.4 11.152 11.262 11.430 11.430 wfi_extrapolate 11 7.9 0.001 0.001 8.434 8.434 cp_fm_cholesky_decompose 22 10.9 8.120 8.239 8.120 8.239 pw_transfer 1547 11.6 0.083 0.091 8.166 8.192 grid_integrate_task_list 128 12.3 7.767 8.147 7.767 8.147 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.777 8.009 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.946 7.979 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.871 7.929 dbcsr_mm_accdrv_process 20762 16.1 2.661 3.447 6.940 7.552 mp_allgather_i34 2507 14.6 2.928 7.190 2.928 7.190 fft_wrap_pw1pw2_140 523 13.2 0.502 0.526 6.950 6.991 density_rs2pw 128 9.7 0.005 0.006 6.294 6.578 calculate_dm_sparse 128 9.5 0.001 0.001 6.127 6.182 fft3d_ps 1291 14.7 2.721 2.797 6.104 6.134 mp_alltoall_d11v 2415 14.1 5.263 5.977 5.263 5.977 dbcsr_complete_redistribute 395 12.7 2.124 2.192 5.563 5.932 grid_collocate_task_list 128 9.7 5.480 5.803 5.480 5.803 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.723 5.733 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.210 5.212 rs_pw_transfer 1046 11.9 0.013 0.013 4.420 4.715 potential_pw2rs 128 12.3 0.026 0.028 4.356 4.369 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.313 4.338 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.570 3.890 multiply_cannon_metrocomm4 7521 15.6 0.023 0.025 1.862 3.880 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.771 3.861 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.558 3.829 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.785 3.813 mp_irecv_dv 28860 15.9 1.828 3.808 1.828 3.808 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.580 3.593 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=175.219000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1430.636364, yerr=54.956632 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_performance_tests/20/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 11.786060E+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 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 2.935435E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144969E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318398E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592338480 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58564. MP_Allreduce 11104 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_ISendRecv 3612 218624. MP_Wait 11682 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 0.023 0.049 296.392 296.392 qs_mol_dyn_low 1 2.0 0.003 0.003 295.861 295.872 qs_forces 11 3.9 0.006 0.006 295.763 295.769 qs_energies 11 4.9 0.020 0.032 286.452 286.468 scf_env_do_scf 11 5.9 0.001 0.001 263.126 263.137 velocity_verlet 10 3.0 0.001 0.001 213.081 213.089 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 139.089 139.091 init_scf_loop 11 6.9 0.000 0.000 123.774 123.778 prepare_preconditioner 11 7.9 0.000 0.000 119.122 119.152 make_preconditioner 11 8.9 0.000 0.000 119.122 119.152 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.280 116.276 qs_scf_new_mos 118 7.6 0.001 0.001 95.542 95.691 qs_scf_loop_do_ot 118 8.6 0.001 0.001 95.541 95.690 ot_scf_mini 118 9.6 0.004 0.004 90.617 90.681 dbcsr_multiply_generic 2529 12.6 0.222 0.241 86.786 87.371 cp_fm_upper_to_full 106 14.8 53.084 75.947 53.084 75.947 multiply_cannon 2529 13.6 0.714 0.755 61.305 62.008 multiply_cannon_loop 2529 14.6 0.475 0.485 57.342 58.797 ot_mini 118 10.6 0.001 0.001 46.035 46.097 dbcsr_complete_redistribute 397 12.7 3.995 4.033 29.876 42.979 copy_fm_to_dbcsr 210 11.7 0.002 0.002 26.301 39.348 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.796 36.676 rebuild_ks_matrix 129 8.3 0.001 0.001 34.942 35.005 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 34.941 35.004 mp_alltoall_i22 720 14.1 21.689 34.945 21.689 34.945 mp_waitall_1 106386 16.7 30.659 34.639 30.659 34.639 cp_fm_cholesky_invert 11 10.9 33.341 33.347 33.341 33.347 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.061 32.115 qs_ot_get_p 129 10.4 0.001 0.001 28.310 28.451 qs_ot_get_derivative 118 11.6 0.002 0.002 24.920 24.970 qs_ot_p2m_diag 84 11.4 0.889 0.894 23.660 23.690 cp_dbcsr_syevd 84 12.4 0.006 0.006 21.893 21.896 multiply_cannon_metrocomm3 10116 15.6 0.024 0.025 20.596 21.831 make_m2s 5058 13.6 0.079 0.081 20.552 21.693 make_images 5058 14.6 3.803 3.908 20.071 21.212 ot_diis_step 118 11.6 0.022 0.023 21.045 21.051 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.059 20.271 apply_single 129 13.6 0.001 0.001 20.059 20.271 cp_fm_diag_elpa 84 13.4 0.000 0.000 18.604 18.605 cp_fm_diag_elpa_base 84 14.4 14.051 15.737 18.600 18.600 multiply_cannon_multrec 10116 15.6 10.632 12.425 18.155 18.266 sum_up_and_integrate 129 10.3 0.323 0.324 16.623 16.717 integrate_v_rspace 129 11.3 0.004 0.004 16.300 16.393 multiply_cannon_sync_h2d 10116 15.6 15.778 15.792 15.778 15.792 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.672 15.696 calculate_rho_elec 129 8.7 0.486 0.487 15.672 15.695 init_scf_run 11 5.9 0.000 0.001 13.235 13.235 scf_env_initial_rho_setup 11 6.9 0.005 0.010 13.234 13.235 make_images_data 5058 15.6 0.061 0.064 11.036 12.979 hybrid_alltoall_any 5245 16.5 1.306 3.047 10.979 12.801 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.839 9.883 wfi_extrapolate 11 7.9 0.001 0.001 9.662 9.662 pw_transfer 1559 11.6 0.091 0.092 9.217 9.221 dbcsr_mm_accdrv_process 20934 16.1 3.767 5.531 7.285 9.147 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.981 8.986 cp_fm_cholesky_decompose 22 10.9 8.789 8.810 8.789 8.810 grid_integrate_task_list 129 12.3 8.599 8.770 8.599 8.770 qs_energies_init_hamiltonians 11 5.9 0.016 0.020 7.986 7.986 mp_alltoall_d11v 2429 14.1 7.750 7.906 7.750 7.906 fft_wrap_pw1pw2_140 527 13.2 0.542 0.545 7.845 7.855 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.341 7.406 fft3d_ps 1301 14.7 2.776 2.789 7.012 7.024 calculate_dm_sparse 129 9.5 0.001 0.001 6.816 6.927 density_rs2pw 129 9.7 0.005 0.005 6.610 6.644 grid_collocate_task_list 129 9.7 6.385 6.419 6.385 6.419 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.300 6.354 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=296.392000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2611.909091, yerr=170.266879 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 1.261400E+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 33251003. MP_ISendRecv 570 19200. MP_Wait 1302 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 0.061 85.355 85.357 qs_energies 1 2.0 0.000 0.001 84.781 84.788 ls_scf 1 3.0 0.004 0.033 83.864 83.872 dbcsr_multiply_generic 111 6.7 0.016 0.017 73.056 73.198 multiply_cannon 111 7.7 0.017 0.020 55.911 57.026 multiply_cannon_loop 111 8.7 0.214 0.231 52.478 53.752 ls_scf_main 1 4.0 0.000 0.002 52.471 52.472 density_matrix_trs4 2 5.0 0.002 0.003 47.023 47.134 ls_scf_init_scf 1 4.0 0.000 0.001 28.316 28.317 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.259 27.299 mp_waitall_1 11316 10.9 22.824 25.996 22.824 25.996 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.153 25.173 multiply_cannon_multrec 2664 9.7 8.190 8.935 15.535 17.261 multiply_cannon_sync_h2d 2664 9.7 13.595 15.418 13.595 15.418 make_m2s 222 7.7 0.009 0.012 13.457 13.947 make_images 222 8.7 0.101 0.111 13.435 13.927 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.726 12.518 make_images_data 222 9.7 0.005 0.006 7.988 8.659 hybrid_alltoall_any 227 10.6 0.218 1.849 6.829 8.044 dbcsr_mm_accdrv_process 4760 10.4 0.516 0.630 6.962 7.894 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.437 7.568 dbcsr_mm_accdrv_process_sort 4760 11.4 6.241 7.125 6.241 7.125 calculate_norms 4752 9.8 5.522 6.218 5.522 6.218 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.050 5.167 mp_sum_l 807 5.4 3.166 4.517 3.166 4.517 make_images_sizes 222 9.7 0.000 0.000 0.863 3.827 mp_alltoall_i44 222 10.7 0.863 3.827 0.863 3.827 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.344 3.312 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.068 3.250 mp_irecv_dv 6231 10.9 2.050 3.224 2.050 3.224 arnoldi_extremal 4 6.8 0.000 0.000 3.188 3.213 arnoldi_normal_ev 4 7.8 0.001 0.003 3.188 3.213 build_subspace 16 8.4 0.009 0.012 3.092 3.095 ls_scf_post 1 4.0 0.000 0.001 3.073 3.084 ls_scf_store_result 1 5.0 0.000 0.000 2.883 2.939 dbcsr_special_finalize 555 9.7 0.005 0.006 2.246 2.710 dbcsr_merge_single_wm 555 10.7 0.456 0.586 2.237 2.701 make_images_pack 222 9.7 2.212 2.641 2.214 2.643 dbcsr_matrix_vector_mult 304 9.0 0.003 0.009 2.305 2.542 dbcsr_matrix_vector_mult_local 304 10.0 2.060 2.446 2.062 2.448 dbcsr_sort_data 658 11.4 2.040 2.434 2.040 2.434 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.203 2.285 buffer_matrices_ensure_size 222 8.7 1.762 2.093 1.762 2.093 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.747 1.749 rebuild_ks_matrix 3 7.3 0.000 0.000 1.737 1.740 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.011 1.737 1.739 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.357000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 2.138079E+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_ISendRecv 282 57600. MP_Wait 828 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 0.062 0.135 92.759 92.760 qs_energies 1 2.0 0.000 0.000 91.602 91.608 ls_scf 1 3.0 0.002 0.014 90.292 90.298 dbcsr_multiply_generic 111 6.7 0.017 0.021 76.088 76.431 multiply_cannon 111 7.7 0.029 0.041 53.409 57.664 ls_scf_main 1 4.0 0.000 0.001 55.930 55.942 multiply_cannon_loop 111 8.7 0.117 0.124 50.095 53.449 density_matrix_trs4 2 5.0 0.002 0.003 49.912 50.185 ls_scf_init_scf 1 4.0 0.000 0.001 30.773 30.774 mp_waitall_1 9246 10.9 21.530 29.977 21.530 29.977 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.641 29.738 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.253 27.267 multiply_cannon_multrec 1332 9.7 13.163 16.584 22.438 26.928 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.829 20.673 make_m2s 222 7.7 0.007 0.008 15.761 16.509 make_images 222 8.7 1.584 1.979 15.731 16.481 dbcsr_mm_accdrv_process 4041 10.4 0.290 0.454 8.879 10.467 make_images_data 222 9.7 0.004 0.005 9.102 10.153 dbcsr_mm_accdrv_process_sort 4041 11.4 8.468 10.013 8.468 10.013 hybrid_alltoall_any 227 10.6 0.523 2.476 8.491 9.591 mp_sum_l 807 5.4 5.637 9.182 5.637 9.182 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.249 7.655 mp_irecv_dv 3311 11.0 3.229 7.596 3.229 7.596 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.257 7.137 calculate_norms 2376 9.8 6.015 6.718 6.015 6.718 multiply_cannon_sync_h2d 1332 9.7 4.763 6.061 4.763 6.061 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.117 5.352 arnoldi_extremal 4 6.8 0.000 0.000 4.695 4.714 arnoldi_normal_ev 4 7.8 0.044 0.079 4.695 4.714 build_subspace 16 8.4 0.014 0.021 4.369 4.372 ls_scf_post 1 4.0 0.001 0.007 3.588 3.592 ls_scf_store_result 1 5.0 0.000 0.000 3.260 3.407 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.129 3.352 dbcsr_matrix_vector_mult_local 304 10.0 2.739 3.217 2.741 3.219 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.795 2.906 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.207 2.717 mp_allgather_i34 111 8.7 0.997 2.492 0.997 2.492 make_images_pack 222 9.7 2.027 2.448 2.029 2.450 dbcsr_sort_data 436 11.2 1.848 2.103 1.848 2.103 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.994 1.996 rebuild_ks_matrix 3 7.3 0.000 0.000 1.976 1.979 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.976 1.979 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.760000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1722.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 2.747060E+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_ISendRecv 186 57600. MP_Wait 732 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 0.052 0.096 95.500 95.502 qs_energies 1 2.0 0.000 0.000 94.749 94.752 ls_scf 1 3.0 0.000 0.002 93.324 93.327 dbcsr_multiply_generic 111 6.7 0.016 0.019 77.909 78.168 ls_scf_main 1 4.0 0.000 0.001 59.031 59.035 multiply_cannon 111 7.7 0.047 0.134 53.511 57.216 multiply_cannon_loop 111 8.7 0.100 0.107 49.956 54.205 density_matrix_trs4 2 5.0 0.002 0.003 53.150 53.344 mp_waitall_1 7374 11.0 24.603 34.107 24.603 34.107 ls_scf_init_scf 1 4.0 0.000 0.001 30.705 30.708 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.571 29.632 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.205 27.218 multiply_cannon_multrec 888 9.7 12.683 15.297 21.246 24.368 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.639 23.126 make_m2s 222 7.7 0.007 0.008 17.417 18.701 make_images 222 8.7 1.981 2.311 17.379 18.661 make_images_data 222 9.7 0.004 0.005 9.993 11.114 hybrid_alltoall_any 227 10.6 0.624 2.876 9.647 11.045 dbcsr_mm_accdrv_process 3754 10.4 0.241 0.417 8.085 9.313 mp_sum_l 807 5.4 5.409 9.257 5.409 9.257 dbcsr_mm_accdrv_process_sort 3754 11.4 7.714 8.896 7.714 8.896 multiply_cannon_sync_h2d 888 9.7 6.019 7.722 6.019 7.722 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.240 7.406 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.499 7.028 mp_irecv_dv 2335 11.1 2.483 6.982 2.483 6.982 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.600 6.505 arnoldi_extremal 4 6.8 0.000 0.000 5.456 5.472 arnoldi_normal_ev 4 7.8 0.001 0.005 5.456 5.472 build_subspace 16 8.4 0.014 0.020 5.155 5.161 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.935 5.122 calculate_norms 1584 9.8 4.331 4.717 4.331 4.717 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.781 4.081 mp_allgather_i34 111 8.7 1.391 3.903 1.391 3.903 ls_scf_post 1 4.0 0.001 0.002 3.587 3.591 dbcsr_matrix_vector_mult_local 304 10.0 3.015 3.587 3.017 3.589 ls_scf_store_result 1 5.0 0.000 0.000 3.313 3.414 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.737 2.832 make_images_sizes 222 9.7 0.000 0.000 1.071 2.189 mp_alltoall_i44 222 10.7 1.071 2.188 1.071 2.188 make_images_pack 222 9.7 1.821 2.139 1.824 2.142 dbcsr_sort_data 325 11.1 1.881 2.138 1.881 2.138 dbcsr_data_release 9322 10.9 1.326 1.921 1.326 1.921 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.502000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2134.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 3.319927E+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_ISendRecv 138 86400. MP_Wait 600 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 0.038 0.071 97.530 97.531 qs_energies 1 2.0 0.000 0.000 96.877 96.884 ls_scf 1 3.0 0.001 0.003 95.228 95.233 dbcsr_multiply_generic 111 6.7 0.018 0.025 79.174 79.403 ls_scf_main 1 4.0 0.000 0.001 59.285 59.286 multiply_cannon 111 7.7 0.053 0.108 51.821 56.450 density_matrix_trs4 2 5.0 0.002 0.003 53.247 53.373 multiply_cannon_loop 111 8.7 0.115 0.127 46.840 50.464 ls_scf_init_scf 1 4.0 0.000 0.001 32.636 32.639 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.506 31.558 mp_waitall_1 6438 11.0 23.063 29.086 23.063 29.086 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.020 29.035 multiply_cannon_multrec 1332 9.7 14.239 17.710 22.162 25.534 make_m2s 222 7.7 0.008 0.009 21.351 22.713 make_images 222 8.7 3.151 3.603 21.300 22.664 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.225 17.498 make_images_data 222 9.7 0.004 0.005 11.959 13.507 hybrid_alltoall_any 227 10.6 0.800 3.780 11.397 13.086 dbcsr_mm_accdrv_process 3641 10.4 0.205 0.413 7.557 9.080 dbcsr_mm_accdrv_process_sort 3641 11.4 7.183 8.665 7.183 8.665 mp_sum_l 807 5.4 4.428 8.454 4.428 8.454 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.441 6.810 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.089 6.112 multiply_cannon_sync_h2d 1332 9.7 5.529 6.067 5.529 6.067 mp_irecv_dv 3229 10.9 2.066 6.023 2.066 6.023 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.710 5.887 arnoldi_extremal 4 6.8 0.000 0.000 5.218 5.232 arnoldi_normal_ev 4 7.8 0.001 0.005 5.218 5.232 build_subspace 16 8.4 0.014 0.021 4.874 4.884 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.527 4.676 calculate_norms 2376 9.8 4.180 4.518 4.180 4.518 mp_allgather_i34 111 8.7 2.095 4.244 2.095 4.244 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.608 3.887 dbcsr_matrix_vector_mult_local 304 10.0 3.207 3.699 3.209 3.701 dbcsr_sort_data 658 11.4 3.126 3.428 3.126 3.428 ls_scf_post 1 4.0 0.000 0.002 3.306 3.313 dbcsr_special_finalize 555 9.7 0.006 0.007 2.886 3.170 dbcsr_merge_single_wm 555 10.7 0.540 0.658 2.877 3.161 ls_scf_store_result 1 5.0 0.000 0.000 3.044 3.109 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.908 2.959 dbcsr_data_release 10477 10.7 1.580 2.416 1.580 2.416 dbcsr_finalize 304 7.8 0.050 0.062 1.802 1.981 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.531000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2723.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 4.636418E+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_ISendRecv 90 115200. MP_Wait 573 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 0.041 0.062 101.663 101.664 qs_energies 1 2.0 0.000 0.000 100.864 100.870 ls_scf 1 3.0 0.000 0.000 98.939 98.945 dbcsr_multiply_generic 111 6.7 0.017 0.019 79.504 79.738 ls_scf_main 1 4.0 0.000 0.000 64.276 64.277 multiply_cannon 111 7.7 0.094 0.201 55.947 60.836 density_matrix_trs4 2 5.0 0.002 0.003 56.800 56.891 multiply_cannon_loop 111 8.7 0.069 0.077 51.446 53.834 mp_waitall_1 5481 11.0 27.872 33.360 27.872 33.360 ls_scf_init_scf 1 4.0 0.000 0.000 31.008 31.013 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.839 29.876 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.594 27.602 multiply_cannon_multrec 444 9.7 14.015 16.079 21.080 23.645 make_m2s 222 7.7 0.006 0.007 18.318 20.752 make_images 222 8.7 3.731 4.433 18.255 20.690 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.949 17.117 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.107 15.114 make_images_data 222 9.7 0.003 0.004 10.539 12.877 hybrid_alltoall_any 227 10.6 0.791 3.788 10.158 12.620 multiply_cannon_sync_h2d 444 9.7 6.518 7.986 6.518 7.986 dbcsr_mm_accdrv_process 3003 10.4 0.176 0.341 6.768 7.887 dbcsr_mm_accdrv_process_sort 3003 11.4 6.441 7.539 6.441 7.539 mp_allgather_i34 111 8.7 2.714 7.133 2.714 7.133 arnoldi_extremal 4 6.8 0.000 0.000 6.570 6.580 arnoldi_normal_ev 4 7.8 0.002 0.005 6.570 6.580 build_subspace 16 8.4 0.015 0.020 6.186 6.195 mp_sum_l 807 5.4 3.293 5.095 3.293 5.095 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.848 5.043 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.689 4.880 dbcsr_matrix_vector_mult_local 304 10.0 3.681 4.148 3.683 4.150 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.861 3.954 make_images_sizes 222 9.7 0.000 0.000 1.108 3.919 mp_alltoall_i44 222 10.7 1.107 3.919 1.107 3.919 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.620 3.882 mp_irecv_dv 1241 11.2 1.602 3.839 1.602 3.839 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.239 3.827 calculate_norms 792 9.8 3.530 3.681 3.530 3.681 ls_scf_post 1 4.0 0.000 0.000 3.655 3.661 ls_scf_store_result 1 5.0 0.000 0.000 3.436 3.478 dbcsr_finalize 304 7.8 0.062 0.079 2.212 2.335 dbcsr_merge_all 275 8.9 0.476 0.530 2.061 2.166 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=101.664000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3596.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a213519a15191f11b61658d664e7104f22d1fbf8_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 8.724857E+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_ISendRecv 84 732600. MP_Wait 309 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 0.095 0.119 110.784 110.786 qs_energies 1 2.0 0.000 0.000 109.070 109.085 ls_scf 1 3.0 0.000 0.000 106.160 106.174 dbcsr_multiply_generic 111 6.7 0.023 0.027 79.550 79.678 ls_scf_main 1 4.0 0.000 0.000 67.375 67.376 density_matrix_trs4 2 5.0 0.002 0.003 58.579 58.644 multiply_cannon 111 7.7 0.176 0.325 50.943 53.277 multiply_cannon_loop 111 8.7 0.067 0.070 47.239 47.813 ls_scf_init_scf 1 4.0 0.000 0.000 35.013 35.014 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.677 33.710 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.402 30.412 mp_waitall_1 4569 11.1 23.765 27.341 23.765 27.341 make_m2s 222 7.7 0.007 0.007 24.735 25.934 make_images 222 8.7 4.589 4.974 24.625 25.822 multiply_cannon_multrec 444 9.7 17.860 18.782 22.468 23.187 hybrid_alltoall_any 227 10.6 1.659 3.634 13.217 16.833 make_images_data 222 9.7 0.003 0.004 13.849 16.592 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.302 11.870 multiply_cannon_sync_h2d 444 9.7 8.846 8.891 8.846 8.891 arnoldi_extremal 4 6.8 0.000 0.000 8.256 8.269 arnoldi_normal_ev 4 7.8 0.003 0.009 8.256 8.269 build_subspace 16 8.4 0.026 0.036 7.707 7.721 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 6.108 6.367 dbcsr_matrix_vector_mult_local 304 10.0 4.997 5.281 5.000 5.283 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.989 5.220 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.779 4.869 dbcsr_mm_accdrv_process 1814 10.4 0.207 0.317 4.445 4.559 dbcsr_mm_accdrv_process_sort 1814 11.4 4.146 4.268 4.146 4.268 make_images_sizes 222 9.7 0.000 0.000 1.545 4.022 mp_alltoall_i44 222 10.7 1.544 4.021 1.544 4.021 mp_allgather_i34 111 8.7 1.309 3.911 1.309 3.911 ls_scf_post 1 4.0 0.000 0.000 3.772 3.786 ls_scf_store_result 1 5.0 0.000 0.000 3.522 3.536 calculate_norms 792 9.8 3.242 3.286 3.242 3.286 dbcsr_finalize 304 7.8 0.082 0.090 3.085 3.132 dbcsr_merge_all 275 8.9 0.891 0.915 2.869 2.908 dbcsr_complete_redistribute 5 7.6 1.445 1.476 2.773 2.903 qs_energies_init_hamiltonians 1 3.0 0.012 0.021 2.880 2.880 dbcsr_data_release 12724 10.6 2.324 2.841 2.324 2.841 matrix_ls_to_qs 2 6.0 0.000 0.000 2.421 2.554 dbcsr_sort_data 325 11.1 2.436 2.499 2.436 2.499 dbcsr_new_transposed 4 7.5 0.243 0.252 2.272 2.285 dbcsr_frobenius_norm 74 6.6 2.053 2.127 2.180 2.219 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=110.786000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6832.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: a213519a15191f11b61658d664e7104f22d1fbf8 Summary: empty Status: OK