=== This is the CP2K Performance-Test === Updating 6a4002a30..5e06bd5b4 Fast-forward src/emd/rt_make_propagators.F | 18 ++++++++---------- src/emd/rt_propagation_methods.F | 10 +++++----- src/emd/rt_propagation_utils.F | 6 +++--- src/emd/rt_propagator_init.F | 2 +- src/motion/rt_propagation.F | 10 ++++------ 5 files changed, 21 insertions(+), 25 deletions(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 5e06bd5b400cb74c2f9aca977c7c2ab82af71875 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.3.2, 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 (12.01.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/01 job id: 44386567 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/02 job id: 44386568 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/03 job id: 44386569 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/04 job id: 44386570 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/05 job id: 44386571 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/06 job id: 44386573 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/07 job id: 44386574 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/08 job id: 44386575 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/09 job id: 44386576 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/10 job id: 44386577 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/11 job id: 44386579 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/12 job id: 44386581 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/13 job id: 44386582 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/14 job id: 44386585 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/15 job id: 44386587 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/16 job id: 44386590 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/17 job id: 44386592 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/18 job id: 44386593 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/19 job id: 44386594 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/20 job id: 44386595 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/21 job id: 44386596 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/22 job id: 44386597 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/23 job id: 44386598 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/24 job id: 44386599 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/25 job id: 44386600 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/26 job id: 44386601 --- 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/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 0.037 0.052 134.705 134.706 farming_run 1 2.0 133.926 133.931 134.659 134.677 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.458180E+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 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 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 0.025 115.926 115.927 qs_energies 1 2.0 0.000 0.000 115.657 115.659 mp2_main 1 3.0 0.000 0.000 113.694 113.697 mp2_gpw_main 1 4.0 0.021 0.030 112.841 112.844 mp2_ri_gpw_compute_in 1 5.0 0.174 0.185 93.866 94.305 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.590 56.032 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.167 41.895 47.299 get_2c_integrals 1 6.0 0.000 0.001 37.474 38.098 integrate_v_rspace 273 8.0 0.439 0.451 25.236 30.449 pw_transfer 6555 10.6 0.374 0.384 27.482 28.234 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 26.129 26.758 grid_integrate_task_list 273 9.0 21.022 26.741 21.022 26.741 fft_wrap_pw1pw2_100 2178 12.4 1.191 1.384 23.643 24.259 compute_2c_integrals 1 7.0 0.002 0.003 19.560 19.566 compute_2c_integrals_loop_lm 1 8.0 0.004 0.008 18.899 19.248 mp2_eri_2c_integrate_gpw 1 9.0 2.369 2.430 18.895 19.246 rpa_ri_compute_en 1 5.0 0.002 0.012 18.867 19.148 cp_fm_cholesky_decompose 12 8.2 17.884 18.553 17.884 18.553 cholesky_decomp 1 7.0 0.000 0.000 16.761 17.435 fft3d_s 5443 13.4 16.212 16.645 16.235 16.668 ao_to_mo_and_store_B_mult_1 272 7.0 10.862 15.588 10.862 15.588 calculate_wavefunction 272 8.0 5.415 5.586 12.572 13.176 rpa_num_int 1 6.0 0.000 0.001 10.605 10.615 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.545 10.615 calc_mat_Q 8 8.0 0.000 0.000 9.370 9.504 contract_S_to_Q 8 9.0 0.000 0.000 8.793 8.926 calc_potential_gpw 544 9.5 0.005 0.006 8.276 8.633 potential_pw2rs 545 10.0 0.107 0.109 7.718 8.495 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.230 8.490 parallel_gemm_fm 14 9.1 0.000 0.000 8.370 8.464 parallel_gemm_fm_cosma 14 10.1 8.370 8.464 8.370 8.464 create_integ_mat 1 6.0 0.014 0.028 7.719 7.728 collocate_single_gaussian 272 10.0 0.039 0.042 7.475 7.676 array2fm 1 7.0 0.000 0.000 6.575 7.134 pw_scatter_s 2720 13.7 4.422 4.619 4.422 4.619 pw_gather_s 2722 13.2 3.889 4.220 3.889 4.220 array2fm_buffer_send 1 8.0 2.918 3.150 2.918 3.150 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.841322, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2729.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.037 395.771 395.773 farming_run 1 2.0 394.812 394.817 395.729 395.733 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.224143E+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 1821 23730. MP_Sync 38 MP_Alltoall 77 2368424. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.045 209.510 209.511 qs_energies 1 2.0 0.000 0.000 209.248 209.262 scf_env_do_scf 1 3.0 0.000 0.000 105.922 105.922 qs_ks_update_qs_env 5 5.0 0.000 0.000 104.980 104.988 rebuild_ks_matrix 4 6.0 0.000 0.000 104.978 104.986 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.066 104.978 104.986 hfx_ks_matrix 4 8.0 0.001 0.001 104.584 104.588 integrate_four_center 4 9.0 0.144 0.463 104.583 104.587 mp2_main 1 3.0 0.000 0.000 103.032 103.047 mp2_gpw_main 1 4.0 0.034 0.053 102.168 102.185 integrate_four_center_main 4 10.0 0.120 0.509 96.583 99.029 integrate_four_center_bin 266 11.0 96.463 98.903 96.463 98.903 init_scf_loop 1 4.0 0.000 0.000 91.745 91.746 mp2_ri_gpw_compute_in 1 5.0 0.065 0.090 74.949 76.123 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.526 55.707 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.162 42.070 47.077 integrate_v_rspace 95 8.0 0.401 0.578 28.447 33.262 pw_transfer 2240 10.6 0.144 0.172 29.959 30.397 fft_wrap_pw1pw2 1868 11.4 0.018 0.023 28.955 29.346 mp2_ri_gpw_compute_en 1 5.0 0.059 0.077 27.068 29.020 grid_integrate_task_list 95 9.0 23.734 28.681 23.734 28.681 ao_to_mo_and_store_B_mult_1 91 7.0 10.770 28.205 10.770 28.205 fft_wrap_pw1pw2_100 730 12.4 1.289 1.498 26.657 27.022 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.832 1.895 25.234 25.243 get_2c_integrals 1 6.0 0.002 0.008 20.311 20.360 compute_2c_integrals 1 7.0 0.003 0.006 19.287 19.296 compute_2c_integrals_loop_lm 1 8.0 0.003 0.008 18.923 19.152 mp2_eri_2c_integrate_gpw 1 9.0 1.728 1.872 18.920 19.151 fft3d_s 1823 13.4 18.446 18.764 18.459 18.777 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.176 14.176 calculate_wavefunction 91 8.0 2.020 2.052 9.744 9.969 mp2_ri_gpw_compute_en_expansio 172 7.0 0.559 0.583 8.800 9.573 potential_pw2rs 186 10.0 0.033 0.036 8.640 9.198 local_gemm 172 8.0 8.241 8.991 8.241 8.991 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.523 8.215 8.747 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.277 8.564 calc_potential_gpw 182 9.5 0.002 0.002 7.945 8.237 collocate_single_gaussian 91 10.0 0.016 0.019 7.913 8.094 mp_sendrecv_dm3 2068 8.0 6.248 6.765 6.248 6.765 mp2_ri_gpw_compute_en_ener 172 7.0 6.343 6.416 6.343 6.416 pw_gather_s 912 13.2 4.889 5.386 4.889 5.386 mp_sync 38 10.4 2.705 5.192 2.705 5.192 pw_scatter_s 910 13.7 3.929 4.250 3.929 4.250 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.177142, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 451.940352E+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 589622. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.082 52.564 52.567 qs_mol_dyn_low 1 2.0 0.003 0.005 52.135 52.143 qs_forces 11 3.9 0.005 0.029 52.052 52.059 qs_energies 11 4.9 0.003 0.019 50.520 50.533 scf_env_do_scf 11 5.9 0.001 0.001 44.369 44.369 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 42.346 42.346 dbcsr_multiply_generic 2286 12.5 0.092 0.096 33.501 33.984 qs_scf_new_mos 108 7.5 0.000 0.001 32.043 32.332 qs_scf_loop_do_ot 108 8.5 0.001 0.001 32.043 32.331 ot_scf_mini 108 9.5 0.002 0.003 30.399 30.578 multiply_cannon 2286 13.5 0.186 0.194 26.029 27.861 multiply_cannon_loop 2286 14.5 1.497 1.562 25.325 27.127 velocity_verlet 10 3.0 0.001 0.002 25.244 25.246 ot_mini 108 10.5 0.001 0.002 19.258 19.489 qs_ot_get_derivative 108 11.5 0.001 0.001 16.321 16.482 mp_waitall_1 245248 16.5 8.315 15.165 8.315 15.165 multiply_cannon_metrocomm3 54864 15.5 0.068 0.074 5.909 13.211 multiply_cannon_multrec 54864 15.5 4.222 6.484 7.753 11.220 rebuild_ks_matrix 119 8.3 0.000 0.000 7.965 8.092 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.965 8.091 multiply_cannon_sync_h2d 54864 15.5 5.926 7.734 5.926 7.734 mp_sum_l 7207 12.9 5.525 7.325 5.525 7.325 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.025 7.141 qs_ot_get_p 119 10.4 0.001 0.001 6.509 6.844 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.338 5.809 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.626 5.728 init_scf_run 11 5.9 0.000 0.001 4.829 4.830 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.829 4.830 dbcsr_mm_accdrv_process 76910 16.1 1.184 1.830 3.452 4.779 sum_up_and_integrate 119 10.3 0.012 0.014 4.545 4.553 integrate_v_rspace 119 11.3 0.002 0.002 4.532 4.542 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.299 4.402 calculate_rho_elec 119 8.7 0.011 0.017 4.299 4.401 qs_ot_p2m_diag 50 11.0 0.005 0.017 3.758 3.833 multiply_cannon_metrocomm1 54864 15.5 0.052 0.057 1.723 3.169 calculate_dm_sparse 119 9.5 0.000 0.000 2.976 3.097 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.884 3.066 apply_single 119 13.6 0.000 0.000 2.883 3.066 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.879 2.880 jit_kernel_multiply 13 15.8 2.205 2.873 2.205 2.873 rs_pw_transfer 974 11.9 0.011 0.013 2.766 2.856 calculate_first_density_matrix 1 7.0 0.000 0.002 2.702 2.705 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.675 2.676 ot_diis_step 108 11.5 0.006 0.008 2.669 2.670 cp_fm_redistribute_end 50 14.0 2.436 2.654 2.441 2.656 cp_fm_diag_elpa_base 50 14.0 0.213 2.562 0.214 2.573 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.443 2.499 density_rs2pw 119 9.7 0.004 0.005 2.299 2.406 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.338 2.340 acc_transpose_blocks 54864 15.5 0.235 0.253 1.818 2.271 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.090 2.155 grid_integrate_task_list 119 12.3 2.035 2.142 2.035 2.142 wfi_extrapolate 11 7.9 0.001 0.001 2.067 2.067 mp_sum_d 4125 12.0 1.369 2.006 1.369 2.006 init_scf_loop 11 6.9 0.000 0.004 2.004 2.004 pw_transfer 1439 11.6 0.053 0.059 1.861 1.931 potential_pw2rs 119 12.3 0.004 0.004 1.881 1.890 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.785 1.858 make_m2s 4572 13.5 0.054 0.056 1.654 1.702 fft3d_ps 1201 14.6 0.372 0.477 1.555 1.622 make_images 4572 14.5 0.132 0.138 1.572 1.621 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.493 1.510 fft_wrap_pw1pw2_140 487 13.2 0.082 0.096 1.375 1.452 mp_waitany 12084 13.8 1.264 1.443 1.264 1.443 mp_alltoall_d11v 2130 13.8 1.288 1.432 1.288 1.432 grid_collocate_task_list 119 9.7 1.296 1.377 1.296 1.377 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.734 1.126 acc_transpose_blocks_kernels 54864 16.5 0.253 0.385 0.826 1.122 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.567000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.423424E+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 1031290. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.029 38.898 38.899 qs_mol_dyn_low 1 2.0 0.003 0.005 38.662 38.669 qs_forces 11 3.9 0.002 0.005 38.325 38.327 qs_energies 11 4.9 0.004 0.024 36.637 36.641 scf_env_do_scf 11 5.9 0.000 0.001 31.419 31.419 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.962 28.962 dbcsr_multiply_generic 2286 12.5 0.113 0.117 21.195 21.570 qs_scf_new_mos 108 7.5 0.001 0.001 19.781 20.030 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.781 20.030 ot_scf_mini 108 9.5 0.002 0.003 18.903 19.079 velocity_verlet 10 3.0 0.001 0.001 18.430 18.437 multiply_cannon 2286 13.5 0.209 0.216 16.238 17.844 multiply_cannon_loop 2286 14.5 0.903 0.974 15.139 16.533 ot_mini 108 10.5 0.001 0.001 11.691 11.927 mp_waitall_1 200699 16.5 5.624 10.884 5.624 10.884 qs_ot_get_derivative 108 11.5 0.001 0.001 9.244 9.423 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.129 9.369 multiply_cannon_multrec 27432 15.5 1.966 4.250 5.862 8.739 rebuild_ks_matrix 119 8.3 0.000 0.000 7.211 7.350 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.211 7.349 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.372 6.497 dbcsr_mm_accdrv_process 47894 16.0 3.094 5.383 3.827 5.667 qs_ot_get_p 119 10.4 0.001 0.001 4.337 4.571 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.583 4.437 sum_up_and_integrate 119 10.3 0.024 0.026 4.200 4.205 integrate_v_rspace 119 11.3 0.002 0.002 4.176 4.183 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.048 4.150 apply_single 119 13.6 0.000 0.000 3.047 4.150 mp_sum_l 7207 12.9 2.027 3.974 2.027 3.974 init_scf_run 11 5.9 0.000 0.001 3.960 3.961 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.960 3.960 qs_rho_update_rho_low 119 7.7 0.001 0.002 3.781 3.808 calculate_rho_elec 119 8.7 0.021 0.024 3.780 3.808 rs_pw_transfer 974 11.9 0.010 0.011 2.574 2.945 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.916 2.935 multiply_cannon_sync_h2d 27432 15.5 2.184 2.869 2.184 2.869 make_m2s 4572 13.5 0.052 0.053 2.469 2.687 make_images 4572 14.5 0.201 0.238 2.380 2.595 density_rs2pw 119 9.7 0.004 0.004 2.105 2.492 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.486 2.487 calculate_first_density_matrix 1 7.0 0.000 0.003 2.450 2.452 init_scf_loop 11 6.9 0.000 0.002 2.434 2.435 ot_diis_step 108 11.5 0.011 0.011 2.397 2.398 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.115 2.208 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.161 2.161 cp_fm_redistribute_end 50 14.0 1.789 2.138 1.793 2.139 calculate_dm_sparse 119 9.5 0.000 0.000 2.021 2.095 cp_fm_diag_elpa_base 50 14.0 0.331 2.031 0.344 2.082 pw_transfer 1439 11.6 0.065 0.074 1.980 2.019 jit_kernel_multiply 10 16.1 0.681 1.970 0.681 1.970 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.956 1.958 grid_integrate_task_list 119 12.3 1.836 1.941 1.836 1.941 potential_pw2rs 119 12.3 0.006 0.006 1.931 1.941 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.888 1.930 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.729 1.769 fft3d_ps 1201 14.6 0.512 0.562 1.586 1.626 prepare_preconditioner 11 7.9 0.000 0.000 1.532 1.559 make_preconditioner 11 8.9 0.000 0.000 1.532 1.559 make_images_data 4572 15.5 0.045 0.051 1.138 1.559 fft_wrap_pw1pw2_140 487 13.2 0.079 0.085 1.489 1.530 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.430 1.486 acc_transpose_blocks 27432 15.5 0.110 0.115 1.186 1.485 wfi_extrapolate 11 7.9 0.001 0.001 1.458 1.458 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.000 1.429 grid_collocate_task_list 119 9.7 1.232 1.359 1.232 1.359 mp_allgather_i34 2286 14.5 0.530 1.285 0.530 1.285 mp_alltoall_d11v 2130 13.8 1.176 1.282 1.176 1.282 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.253 1.261 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.155 1.203 mp_sum_d 4125 12.0 0.607 1.076 0.607 1.076 qs_energies_init_hamiltonians 11 5.9 0.008 0.019 0.968 0.981 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.147 0.557 0.934 mp_waitany 5720 13.7 0.533 0.925 0.533 0.925 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.897 0.910 acc_transpose_blocks_kernels 27432 16.5 0.182 0.273 0.651 0.859 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.607 0.754 0.814 mp_alltoall_z22v 1201 16.6 0.685 0.786 0.685 0.786 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.899000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 522.022912E+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_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 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 0.012 0.031 32.373 32.375 qs_mol_dyn_low 1 2.0 0.003 0.004 32.008 32.016 qs_forces 11 3.9 0.002 0.002 31.947 31.948 qs_energies 11 4.9 0.002 0.002 30.368 30.371 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 23.028 23.028 dbcsr_multiply_generic 2286 12.5 0.102 0.111 16.385 16.476 velocity_verlet 10 3.0 0.001 0.001 15.328 15.330 qs_scf_new_mos 108 7.5 0.001 0.001 14.747 14.766 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.747 14.765 ot_scf_mini 108 9.5 0.002 0.002 14.026 14.046 multiply_cannon 2286 13.5 0.196 0.201 13.114 13.910 multiply_cannon_loop 2286 14.5 0.638 0.669 12.337 13.160 ot_mini 108 10.5 0.001 0.001 8.677 8.693 multiply_cannon_multrec 18288 15.5 1.962 2.914 6.934 7.221 qs_ot_get_derivative 108 11.5 0.001 0.001 7.175 7.195 rebuild_ks_matrix 119 8.3 0.000 0.000 6.419 6.438 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.419 6.437 dbcsr_mm_accdrv_process 38222 16.0 4.065 5.516 4.887 5.855 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.669 5.686 sum_up_and_integrate 119 10.3 0.031 0.032 3.994 3.998 integrate_v_rspace 119 11.3 0.002 0.003 3.963 3.972 init_scf_run 11 5.9 0.000 0.001 3.574 3.574 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.573 3.573 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.507 3.518 calculate_rho_elec 119 8.7 0.030 0.031 3.507 3.518 mp_waitall_1 158411 16.6 2.524 3.384 2.524 3.384 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.666 3.306 qs_ot_get_p 119 10.4 0.001 0.001 3.209 3.234 init_scf_loop 11 6.9 0.000 0.000 2.568 2.568 rs_pw_transfer 974 11.9 0.009 0.010 2.284 2.536 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.054 2.360 apply_single 119 13.6 0.000 0.000 2.054 2.360 calculate_first_density_matrix 1 7.0 0.000 0.000 2.347 2.348 density_rs2pw 119 9.7 0.004 0.004 2.024 2.285 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.416 2.153 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.128 2.134 jit_kernel_multiply 11 16.1 0.771 2.130 0.771 2.130 pw_transfer 1439 11.6 0.066 0.071 1.989 2.002 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.895 1.913 make_m2s 4572 13.5 0.044 0.045 1.784 1.906 calculate_dm_sparse 119 9.5 0.000 0.000 1.880 1.894 grid_integrate_task_list 119 12.3 1.799 1.885 1.799 1.885 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.866 1.867 make_images 4572 14.5 0.190 0.202 1.699 1.820 potential_pw2rs 119 12.3 0.007 0.008 1.764 1.770 prepare_preconditioner 11 7.9 0.000 0.000 1.753 1.756 make_preconditioner 11 8.9 0.000 0.000 1.753 1.756 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.739 1.741 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.607 1.690 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.645 1.646 cp_fm_redistribute_end 50 14.0 1.226 1.625 1.227 1.625 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.597 1.603 cp_fm_diag_elpa_base 50 14.0 0.382 1.547 0.396 1.594 fft3d_ps 1201 14.6 0.524 0.542 1.575 1.594 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.548 1.560 fft_wrap_pw1pw2_140 487 13.2 0.090 0.094 1.514 1.528 multiply_cannon_sync_h2d 18288 15.5 1.372 1.519 1.372 1.519 ot_diis_step 108 11.5 0.011 0.011 1.484 1.484 mp_sum_l 7207 12.9 1.094 1.402 1.094 1.402 grid_collocate_task_list 119 9.7 1.215 1.385 1.215 1.385 acc_transpose_blocks 18288 15.5 0.076 0.077 1.284 1.305 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.250 1.255 wfi_extrapolate 11 7.9 0.001 0.001 1.181 1.181 make_images_data 4572 15.5 0.045 0.049 0.808 0.946 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.933 0.935 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.881 0.904 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.690 0.866 acc_transpose_blocks_kernels 18288 16.5 0.210 0.219 0.843 0.857 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.821 0.822 mp_alltoall_z22v 1201 16.6 0.653 0.762 0.653 0.762 mp_waitany 9880 13.7 0.501 0.761 0.501 0.761 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.122 0.495 0.741 cp_fm_cholesky_invert 11 10.9 0.736 0.740 0.736 0.740 mp_alltoall_d11v 2130 13.8 0.660 0.733 0.660 0.733 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.653 0.727 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.375000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.000000, yerr=1.705606 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 554.098688E+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_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 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 0.070 0.125 35.534 35.538 qs_mol_dyn_low 1 2.0 0.003 0.004 35.189 35.197 qs_forces 11 3.9 0.002 0.006 35.130 35.132 qs_energies 11 4.9 0.003 0.007 33.399 33.405 scf_env_do_scf 11 5.9 0.001 0.003 28.216 28.218 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 24.726 24.727 dbcsr_multiply_generic 2286 12.5 0.099 0.103 18.124 18.257 velocity_verlet 10 3.0 0.001 0.001 18.017 18.019 qs_scf_new_mos 108 7.5 0.001 0.001 16.076 16.131 qs_scf_loop_do_ot 108 8.5 0.002 0.009 16.076 16.130 ot_scf_mini 108 9.5 0.009 0.023 15.145 15.202 multiply_cannon 2286 13.5 0.230 0.268 14.413 14.835 multiply_cannon_loop 2286 14.5 0.942 0.977 13.451 13.913 ot_mini 108 10.5 0.001 0.001 9.224 9.293 multiply_cannon_multrec 27432 15.5 2.327 2.972 8.630 8.953 dbcsr_mm_accdrv_process 47916 15.9 5.374 7.181 6.207 7.475 qs_ot_get_derivative 108 11.5 0.001 0.001 7.411 7.469 rebuild_ks_matrix 119 8.3 0.000 0.000 6.707 6.751 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 6.706 6.750 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.963 6.001 sum_up_and_integrate 119 10.3 0.036 0.039 3.897 3.904 integrate_v_rspace 119 11.3 0.002 0.002 3.861 3.869 init_scf_run 11 5.9 0.000 0.002 3.802 3.802 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.801 3.802 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.631 3.666 calculate_rho_elec 119 8.7 0.040 0.046 3.631 3.666 init_scf_loop 11 6.9 0.001 0.003 3.468 3.469 qs_ot_get_p 119 10.4 0.001 0.001 3.293 3.373 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.702 3.181 prepare_preconditioner 11 7.9 0.000 0.000 2.610 2.617 make_preconditioner 11 8.9 0.000 0.001 2.610 2.617 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.111 2.565 apply_single 119 13.6 0.000 0.000 2.111 2.565 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.215 2.541 calculate_first_density_matrix 1 7.0 0.001 0.003 2.414 2.415 mp_waitall_1 137007 16.6 1.824 2.402 1.824 2.402 make_m2s 4572 13.5 0.054 0.055 2.224 2.328 rs_pw_transfer 974 11.9 0.009 0.009 2.095 2.289 density_rs2pw 119 9.7 0.004 0.004 2.047 2.232 make_images 4572 14.5 0.271 0.333 2.117 2.220 pw_transfer 1439 11.6 0.066 0.071 2.152 2.190 calculate_dm_sparse 119 9.5 0.000 0.000 2.128 2.178 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.058 2.102 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.019 2.029 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.940 1.970 grid_integrate_task_list 119 12.3 1.821 1.902 1.821 1.902 jit_kernel_multiply 10 16.0 0.773 1.847 0.773 1.847 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.835 1.836 fft3d_ps 1201 14.6 0.560 0.613 1.732 1.774 ot_diis_step 108 11.5 0.012 0.012 1.773 1.774 fft_wrap_pw1pw2_140 487 13.2 0.088 0.096 1.690 1.737 mp_sum_l 7207 12.9 1.102 1.699 1.102 1.699 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.695 1.696 potential_pw2rs 119 12.3 0.008 0.010 1.651 1.655 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.620 1.631 acc_transpose_blocks 27432 15.5 0.113 0.116 1.446 1.462 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.446 1.447 cp_fm_redistribute_end 50 14.0 0.957 1.422 0.959 1.422 cp_fm_diag_elpa_base 50 14.0 0.442 1.356 0.461 1.395 grid_collocate_task_list 119 9.7 1.222 1.343 1.222 1.343 wfi_extrapolate 11 7.9 0.001 0.001 1.337 1.337 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.318 1.327 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.722 1.268 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.110 1.131 dbcsr_complete_redistribute 329 12.2 0.130 0.167 0.842 1.116 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.112 1.114 cp_fm_upper_to_full 72 13.5 0.800 1.111 0.800 1.111 multiply_cannon_sync_h2d 27432 15.5 0.985 1.082 0.985 1.082 make_images_data 4572 15.5 0.045 0.048 0.870 0.985 mp_alltoall_d11v 2130 13.8 0.750 0.963 0.750 0.963 hybrid_alltoall_any 4725 16.4 0.062 0.152 0.752 0.927 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.610 0.878 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.875 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.835 0.839 acc_transpose_blocks_kernels 27432 16.5 0.268 0.276 0.819 0.828 mp_alltoall_z22v 1201 16.6 0.756 0.797 0.756 0.797 cp_fm_cholesky_invert 11 10.9 0.793 0.796 0.793 0.796 mp_alltoall_i22 627 13.8 0.430 0.728 0.430 0.728 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.538000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.454545, yerr=3.701686 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 606.355456E+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_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 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 0.012 0.031 29.280 29.281 qs_mol_dyn_low 1 2.0 0.003 0.003 29.050 29.059 qs_forces 11 3.9 0.002 0.002 28.469 28.470 qs_energies 11 4.9 0.001 0.001 26.751 26.753 scf_env_do_scf 11 5.9 0.000 0.001 21.653 21.654 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.118 19.119 velocity_verlet 10 3.0 0.001 0.001 15.087 15.122 dbcsr_multiply_generic 2286 12.5 0.092 0.094 12.522 12.592 qs_scf_new_mos 108 7.5 0.001 0.001 10.988 11.011 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.988 11.010 multiply_cannon 2286 13.5 0.231 0.240 9.925 10.402 ot_scf_mini 108 9.5 0.002 0.002 10.314 10.338 multiply_cannon_loop 2286 14.5 0.331 0.343 9.009 9.201 rebuild_ks_matrix 119 8.3 0.000 0.000 6.080 6.098 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.080 6.098 multiply_cannon_multrec 9144 15.5 1.598 1.852 5.769 6.002 ot_mini 108 10.5 0.001 0.001 5.721 5.749 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.424 5.439 qs_ot_get_derivative 108 11.5 0.001 0.001 4.423 4.448 dbcsr_mm_accdrv_process 12550 15.8 3.207 3.918 4.069 4.139 sum_up_and_integrate 119 10.3 0.038 0.042 3.732 3.737 integrate_v_rspace 119 11.3 0.002 0.003 3.694 3.699 init_scf_run 11 5.9 0.000 0.001 3.632 3.632 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.632 3.632 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.595 3.602 calculate_rho_elec 119 8.7 0.060 0.061 3.595 3.602 qs_ot_get_p 119 10.4 0.001 0.001 2.800 2.841 init_scf_loop 11 6.9 0.000 0.000 2.515 2.515 calculate_first_density_matrix 1 7.0 0.000 0.000 2.462 2.462 pw_transfer 1439 11.6 0.066 0.069 2.206 2.216 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.112 2.124 density_rs2pw 119 9.7 0.004 0.004 1.939 2.080 mp_waitall_1 115863 16.7 1.487 2.030 1.487 2.030 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.936 1.938 grid_integrate_task_list 119 12.3 1.853 1.924 1.853 1.924 make_m2s 4572 13.5 0.034 0.035 1.772 1.920 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.871 1.873 rs_pw_transfer 974 11.9 0.008 0.008 1.701 1.842 make_images 4572 14.5 0.268 0.302 1.683 1.830 fft3d_ps 1201 14.6 0.564 0.575 1.772 1.785 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.771 1.779 jit_kernel_multiply 10 15.8 0.823 1.764 0.823 1.764 prepare_preconditioner 11 7.9 0.000 0.000 1.760 1.763 make_preconditioner 11 8.9 0.000 0.000 1.760 1.763 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.732 1.743 calculate_dm_sparse 119 9.5 0.000 0.000 1.694 1.714 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.648 1.673 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.669 1.669 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.521 1.534 potential_pw2rs 119 12.3 0.010 0.010 1.455 1.459 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.394 1.395 grid_collocate_task_list 119 9.7 1.271 1.387 1.271 1.387 cp_fm_redistribute_end 50 14.0 0.695 1.370 0.696 1.371 cp_fm_diag_elpa_base 50 14.0 0.629 1.296 0.673 1.353 ot_diis_step 108 11.5 0.012 0.013 1.285 1.285 acc_transpose_blocks 9144 15.5 0.038 0.039 1.277 1.285 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.221 1.222 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.163 1.180 apply_single 119 13.6 0.000 0.000 1.162 1.180 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.149 1.160 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.115 1.121 wfi_extrapolate 11 7.9 0.001 0.001 1.120 1.120 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 1.036 1.041 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.797 1.028 make_images_data 4572 15.5 0.039 0.042 0.836 1.010 jit_kernel_transpose 5 15.6 0.920 0.922 0.920 0.922 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.866 0.917 cp_fm_cholesky_invert 11 10.9 0.900 0.903 0.900 0.903 mp_alltoall_d11v 2130 13.8 0.751 0.868 0.751 0.868 multiply_cannon_sync_h2d 9144 15.5 0.716 0.787 0.716 0.787 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.780 0.782 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.726 0.738 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.675 0.727 mp_alltoall_z22v 1201 16.6 0.655 0.715 0.655 0.715 mp_allgather_i34 2286 14.5 0.251 0.689 0.251 0.689 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.358 0.670 yz_to_x 606 15.1 0.265 0.277 0.606 0.626 mp_sum_l 7207 12.9 0.400 0.616 0.400 0.616 x_to_yz 595 16.2 0.276 0.288 0.589 0.598 mp_sum_dm 438 4.9 0.549 0.589 0.549 0.589 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.281000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=575.545455, yerr=3.525585 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 742.555648E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.075 43.003 43.005 qs_mol_dyn_low 1 2.0 0.003 0.004 42.741 42.748 qs_forces 11 3.9 0.002 0.002 42.677 42.678 qs_energies 11 4.9 0.001 0.001 40.654 40.657 scf_env_do_scf 11 5.9 0.001 0.001 34.871 34.871 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.685 26.687 velocity_verlet 10 3.0 0.001 0.001 24.262 24.267 dbcsr_multiply_generic 2286 12.5 0.100 0.100 17.869 18.050 qs_scf_new_mos 108 7.5 0.001 0.001 16.502 16.584 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.501 16.583 ot_scf_mini 108 9.5 0.002 0.002 15.398 15.494 multiply_cannon 2286 13.5 0.303 0.313 13.795 14.702 multiply_cannon_loop 2286 14.5 0.344 0.349 12.464 13.389 ot_mini 108 10.5 0.001 0.001 9.135 9.249 multiply_cannon_multrec 9144 15.5 3.408 4.718 8.610 8.709 init_scf_loop 11 6.9 0.000 0.000 8.156 8.160 rebuild_ks_matrix 119 8.3 0.000 0.000 7.614 7.748 qs_ks_build_kohn_sham_matrix 119 9.3 0.048 0.050 7.614 7.747 qs_ot_get_derivative 108 11.5 0.001 0.001 7.073 7.166 prepare_preconditioner 11 7.9 0.000 0.000 7.112 7.129 make_preconditioner 11 8.9 0.000 0.000 7.112 7.129 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.893 7.013 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.632 7.002 dbcsr_mm_accdrv_process 12550 15.8 4.218 5.933 5.075 6.360 cp_fm_upper_to_full 72 14.2 3.204 4.621 3.204 4.621 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.437 4.446 calculate_rho_elec 119 8.7 0.118 0.121 4.436 4.445 sum_up_and_integrate 119 10.3 0.065 0.067 4.185 4.191 integrate_v_rspace 119 11.3 0.003 0.003 4.119 4.124 init_scf_run 11 5.9 0.000 0.001 3.729 3.730 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.729 3.729 qs_ot_get_p 119 10.4 0.001 0.001 3.472 3.602 mp_waitall_1 94719 16.7 2.622 3.557 2.622 3.557 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.596 2.995 pw_transfer 1439 11.6 0.068 0.069 2.957 2.965 dbcsr_complete_redistribute 329 12.2 0.286 0.291 2.064 2.914 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.858 2.866 make_m2s 4572 13.5 0.038 0.038 2.452 2.680 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.730 2.582 make_images 4572 14.5 0.353 0.386 2.332 2.559 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.202 2.520 apply_single 119 13.6 0.000 0.000 2.202 2.520 fft3d_ps 1201 14.6 0.595 0.604 2.484 2.492 fft_wrap_pw1pw2_140 487 13.2 0.094 0.097 2.383 2.391 density_rs2pw 119 9.7 0.004 0.004 2.339 2.362 mp_alltoall_i22 627 13.8 1.515 2.347 1.515 2.347 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.475 2.320 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.418 2.245 calculate_first_density_matrix 1 7.0 0.000 0.000 2.198 2.202 calculate_dm_sparse 119 9.5 0.000 0.000 2.169 2.191 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.148 2.149 grid_integrate_task_list 119 12.3 2.061 2.077 2.061 2.077 ot_diis_step 108 11.5 0.014 0.014 2.033 2.033 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.960 1.960 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.859 1.859 mp_sum_l 7207 12.9 1.097 1.817 1.097 1.817 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.765 1.811 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.771 1.772 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.619 1.649 cp_fm_cholesky_invert 11 10.9 1.624 1.628 1.624 1.628 rs_pw_transfer 974 11.9 0.009 0.009 1.582 1.610 potential_pw2rs 119 12.3 0.014 0.015 1.559 1.563 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.535 1.536 cp_fm_diag_elpa_base 50 14.0 1.387 1.442 1.533 1.534 grid_collocate_task_list 119 9.7 1.492 1.505 1.492 1.505 make_images_data 4572 15.5 0.043 0.046 1.176 1.480 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.193 1.477 wfi_extrapolate 11 7.9 0.001 0.001 1.461 1.461 jit_kernel_multiply 8 15.6 0.830 1.399 0.830 1.399 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.221 1.230 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.172 1.194 mp_alltoall_d11v 2130 13.8 1.170 1.190 1.170 1.190 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.109 1.123 multiply_cannon_sync_h2d 9144 15.5 1.043 1.047 1.043 1.047 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.993 1.038 yz_to_x 606 15.1 0.459 0.470 1.014 1.030 mp_alltoall_z22v 1201 16.6 0.937 0.986 0.937 0.986 qs_create_task_list 11 7.9 0.001 0.001 0.945 0.955 generate_qs_task_list 11 8.9 0.371 0.389 0.944 0.955 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.937 0.950 acc_transpose_blocks 9144 15.5 0.038 0.038 0.894 0.899 x_to_yz 595 16.2 0.479 0.490 0.861 0.867 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.005000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=696.272727, yerr=14.359608 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 502.112256E+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 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 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 0.023 0.059 84.559 84.568 qs_mol_dyn_low 1 2.0 0.003 0.004 84.238 84.248 qs_forces 11 3.9 0.002 0.002 83.502 83.502 qs_energies 11 4.9 0.001 0.001 80.629 80.643 scf_env_do_scf 11 5.9 0.000 0.001 71.767 71.770 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 66.108 66.109 dbcsr_multiply_generic 2055 12.4 0.106 0.110 52.282 52.516 qs_scf_new_mos 99 7.5 0.000 0.001 48.563 48.679 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.562 48.679 ot_scf_mini 99 9.5 0.002 0.002 46.166 46.258 multiply_cannon 2055 13.4 0.179 0.185 42.664 43.738 velocity_verlet 10 3.0 0.001 0.001 43.209 43.216 multiply_cannon_loop 2055 14.4 1.544 1.604 41.698 42.814 ot_mini 99 10.5 0.001 0.001 27.929 28.070 qs_ot_get_derivative 99 11.5 0.001 0.001 21.036 21.177 multiply_cannon_multrec 49320 15.4 12.097 12.836 17.119 17.898 rebuild_ks_matrix 110 8.3 0.000 0.001 14.714 14.849 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.713 14.849 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.903 13.025 mp_waitall_1 220248 16.4 11.762 12.831 11.762 12.831 multiply_cannon_sync_h2d 49320 15.4 9.909 10.641 9.909 10.641 qs_ot_get_p 110 10.4 0.001 0.001 9.731 9.908 multiply_cannon_metrocomm3 49320 15.4 0.080 0.084 7.143 8.569 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.209 7.706 apply_single 110 13.6 0.000 0.001 7.209 7.706 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.119 7.684 sum_up_and_integrate 110 10.3 0.036 0.043 7.196 7.211 integrate_v_rspace 110 11.3 0.002 0.003 7.159 7.182 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.757 6.837 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.667 6.816 calculate_rho_elec 110 8.6 0.021 0.026 6.667 6.815 init_scf_run 11 5.9 0.000 0.001 6.811 6.811 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.811 6.811 ot_diis_step 99 11.5 0.005 0.006 6.555 6.555 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.434 6.462 mp_sum_l 6514 12.8 5.337 6.285 5.337 6.285 init_scf_loop 11 6.9 0.000 0.000 5.631 5.632 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.401 5.401 dbcsr_mm_accdrv_process 87628 16.1 2.079 2.182 4.902 5.194 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.907 4.908 cp_fm_redistribute_end 48 14.0 4.283 4.882 4.287 4.883 cp_fm_diag_elpa_base 48 14.0 0.590 4.744 0.594 4.776 rs_pw_transfer 902 11.9 0.011 0.013 3.742 4.350 density_rs2pw 110 9.6 0.004 0.005 3.468 4.026 wfi_extrapolate 11 7.9 0.001 0.001 4.020 4.020 make_m2s 4110 13.4 0.060 0.065 3.838 3.951 calculate_dm_sparse 110 9.5 0.000 0.001 3.771 3.874 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.738 3.865 make_images 4110 14.4 0.177 0.189 3.743 3.860 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.505 3.509 prepare_preconditioner 11 7.9 0.000 0.000 3.427 3.448 make_preconditioner 11 8.9 0.000 0.000 3.427 3.448 grid_integrate_task_list 110 12.3 3.232 3.396 3.232 3.396 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.287 3.385 pw_transfer 1331 11.6 0.055 0.066 3.296 3.361 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.208 3.276 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.217 3.265 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.195 3.242 fft_wrap_pw1pw2_140 451 13.1 0.171 0.192 2.724 2.799 fft3d_ps 1111 14.6 0.783 0.883 2.736 2.799 potential_pw2rs 110 12.3 0.006 0.007 2.722 2.745 calculate_first_density_matrix 1 7.0 0.000 0.000 2.696 2.706 jit_kernel_multiply 13 15.9 2.538 2.557 2.538 2.557 mp_alltoall_d11v 2046 13.8 2.099 2.529 2.099 2.529 mp_waitany 14300 13.8 1.815 2.306 1.815 2.306 grid_collocate_task_list 110 9.6 2.084 2.302 2.084 2.302 acc_transpose_blocks 49320 15.4 0.226 0.235 2.179 2.258 mp_sum_d 3879 11.9 1.514 2.049 1.514 2.049 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.951 1.969 make_images_data 4110 15.4 0.043 0.045 1.750 1.894 cp_fm_cholesky_invert 11 10.9 1.856 1.860 1.856 1.860 hybrid_alltoall_any 4261 16.3 0.082 0.484 1.537 1.841 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.568000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.818182, yerr=2.622219 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 585.076736E+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 3309696. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 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 0.017 0.032 70.200 70.202 qs_mol_dyn_low 1 2.0 0.003 0.005 69.898 69.908 qs_forces 11 3.9 0.003 0.009 69.512 69.514 qs_energies 11 4.9 0.002 0.006 66.150 66.154 scf_env_do_scf 11 5.9 0.001 0.006 57.360 57.364 scf_env_do_scf_inner_loop 99 6.5 0.003 0.012 49.560 49.561 dbcsr_multiply_generic 2055 12.4 0.122 0.132 37.855 38.016 velocity_verlet 10 3.0 0.001 0.001 36.963 36.971 qs_scf_new_mos 99 7.5 0.001 0.001 32.987 33.118 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.987 33.117 multiply_cannon 2055 13.4 0.222 0.244 31.056 32.170 ot_scf_mini 99 9.5 0.003 0.003 31.312 31.435 multiply_cannon_loop 2055 14.4 0.928 0.952 29.728 30.771 ot_mini 99 10.5 0.001 0.001 18.387 18.511 multiply_cannon_multrec 24660 15.4 7.594 9.565 13.784 15.951 rebuild_ks_matrix 110 8.3 0.000 0.001 13.990 14.128 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 13.990 14.128 qs_ot_get_derivative 99 11.5 0.001 0.001 12.536 12.662 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.327 12.446 mp_waitall_1 176588 16.5 7.791 10.832 7.791 10.832 multiply_cannon_sync_h2d 24660 15.4 7.021 8.177 7.021 8.177 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.224 7.944 init_scf_loop 11 6.9 0.001 0.003 7.764 7.764 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.552 7.267 apply_single 110 13.6 0.000 0.001 6.552 7.267 sum_up_and_integrate 110 10.3 0.052 0.059 6.755 6.764 integrate_v_rspace 110 11.3 0.002 0.002 6.703 6.715 qs_ot_get_p 110 10.4 0.001 0.001 6.279 6.434 init_scf_run 11 5.9 0.000 0.001 6.371 6.372 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.371 6.372 qs_rho_update_rho_low 110 7.6 0.006 0.049 6.349 6.362 calculate_rho_elec 110 8.6 0.040 0.049 6.343 6.361 dbcsr_mm_accdrv_process 52282 16.1 4.678 5.825 6.031 6.297 ot_diis_step 99 11.5 0.010 0.010 5.796 5.797 prepare_preconditioner 11 7.9 0.000 0.000 5.663 5.687 make_preconditioner 11 8.9 0.000 0.000 5.663 5.687 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.740 5.474 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.252 5.406 make_m2s 4110 13.4 0.057 0.059 4.288 4.771 make_images 4110 14.4 0.401 0.445 4.179 4.659 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.311 4.331 pw_transfer 1331 11.6 0.066 0.075 3.847 3.996 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.740 3.890 density_rs2pw 110 9.6 0.004 0.005 3.430 3.876 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.868 3.868 rs_pw_transfer 902 11.9 0.012 0.014 3.090 3.580 wfi_extrapolate 11 7.9 0.001 0.001 3.556 3.556 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.390 3.392 grid_integrate_task_list 110 12.3 3.143 3.354 3.143 3.354 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.225 3.300 fft_wrap_pw1pw2_140 451 13.1 0.202 0.220 3.138 3.298 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.281 3.283 cp_fm_redistribute_end 48 14.0 2.452 3.249 2.455 3.251 fft3d_ps 1111 14.6 1.102 1.339 3.089 3.228 cp_fm_diag_elpa_base 48 14.0 0.761 3.096 0.792 3.179 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.975 3.019 calculate_dm_sparse 110 9.5 0.001 0.001 2.975 3.008 make_images_data 4110 15.4 0.047 0.050 2.380 2.860 hybrid_alltoall_any 4261 16.3 0.103 0.442 2.072 2.839 calculate_first_density_matrix 1 7.0 0.000 0.002 2.666 2.670 cp_fm_cholesky_invert 11 10.9 2.637 2.644 2.637 2.644 mp_sum_l 6514 12.8 1.885 2.583 1.885 2.583 potential_pw2rs 110 12.3 0.008 0.009 2.561 2.572 grid_collocate_task_list 110 9.6 2.048 2.483 2.048 2.483 jit_kernel_multiply 12 16.2 0.999 2.304 0.999 2.304 mp_alltoall_d11v 2046 13.8 1.807 2.108 1.807 2.108 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.970 1.989 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.893 1.895 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.830 1.846 mp_waitany 10164 13.8 1.225 1.739 1.225 1.739 multiply_cannon_metrocomm4 22605 15.4 0.072 0.075 0.798 1.713 mp_allgather_i34 2055 14.4 0.657 1.675 0.657 1.675 mp_irecv_dv 57340 16.2 0.675 1.597 0.675 1.597 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.533 1.548 acc_transpose_blocks 24660 15.4 0.111 0.115 1.522 1.544 dbcsr_complete_redistribute 325 12.2 0.288 0.366 1.236 1.503 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.369 1.479 rs_pw_transfer_RS2PW_140 121 11.5 0.206 0.215 0.970 1.452 mp_alltoall_z22v 1111 16.6 1.298 1.415 1.298 1.415 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.202000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=553.909091, yerr=6.171114 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 658.698240E+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_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 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 0.070 0.128 61.531 61.533 qs_mol_dyn_low 1 2.0 0.004 0.005 61.004 61.015 qs_forces 11 3.9 0.004 0.012 60.937 60.939 qs_energies 11 4.9 0.003 0.007 57.610 57.614 scf_env_do_scf 11 5.9 0.001 0.001 49.162 49.162 scf_env_do_scf_inner_loop 99 6.5 0.009 0.018 40.487 40.488 velocity_verlet 10 3.0 0.001 0.002 32.900 32.902 dbcsr_multiply_generic 2055 12.4 0.109 0.111 28.767 29.052 qs_scf_new_mos 99 7.5 0.001 0.001 25.170 25.289 qs_scf_loop_do_ot 99 8.5 0.002 0.010 25.169 25.288 ot_scf_mini 99 9.5 0.007 0.032 23.940 24.075 multiply_cannon 2055 13.4 0.212 0.225 22.412 23.622 multiply_cannon_loop 2055 14.4 0.616 0.633 21.261 22.440 ot_mini 99 10.5 0.001 0.001 13.622 13.760 rebuild_ks_matrix 110 8.3 0.000 0.001 12.728 12.891 qs_ks_build_kohn_sham_matrix 110 9.3 0.016 0.045 12.728 12.890 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.180 11.328 mp_waitall_1 139946 16.5 7.339 10.698 7.339 10.698 multiply_cannon_multrec 16440 15.4 3.725 4.746 9.721 10.657 qs_ot_get_derivative 99 11.5 0.001 0.001 9.113 9.252 init_scf_loop 11 6.9 0.003 0.010 8.640 8.644 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.648 7.901 prepare_preconditioner 11 7.9 0.000 0.000 6.757 6.778 make_preconditioner 11 8.9 0.000 0.001 6.757 6.778 sum_up_and_integrate 110 10.3 0.060 0.061 6.731 6.773 integrate_v_rspace 110 11.3 0.002 0.003 6.671 6.713 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.111 6.469 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.108 6.123 calculate_rho_elec 110 8.6 0.058 0.059 6.107 6.123 dbcsr_mm_accdrv_process 34862 16.1 4.919 5.486 5.850 6.033 init_scf_run 11 5.9 0.000 0.001 5.918 5.919 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.918 5.918 qs_ot_get_p 110 10.4 0.001 0.002 5.498 5.689 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.022 5.452 apply_single 110 13.6 0.000 0.000 5.022 5.452 make_m2s 4110 13.4 0.049 0.051 4.206 4.532 ot_diis_step 99 11.5 0.010 0.011 4.480 4.480 make_images 4110 14.4 0.395 0.513 4.092 4.417 density_rs2pw 110 9.6 0.004 0.005 3.187 4.404 multiply_cannon_sync_h2d 16440 15.4 3.622 4.146 3.622 4.146 rs_pw_transfer 902 11.9 0.010 0.011 2.790 3.982 pw_transfer 1331 11.6 0.066 0.074 3.837 3.878 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.871 3.876 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.166 3.808 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.730 3.773 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.531 3.532 grid_integrate_task_list 110 12.3 3.169 3.451 3.169 3.451 fft_wrap_pw1pw2_140 451 13.1 0.213 0.217 3.220 3.265 fft3d_ps 1111 14.6 1.087 1.097 3.029 3.072 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.008 3.010 wfi_extrapolate 11 7.9 0.001 0.001 2.992 2.993 cp_fm_redistribute_end 48 14.0 1.877 2.981 1.880 2.982 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.965 2.970 cp_fm_diag_elpa_base 48 14.0 1.038 2.826 1.097 2.938 make_images_data 4110 15.4 0.043 0.046 2.459 2.885 calculate_first_density_matrix 1 7.0 0.001 0.002 2.807 2.808 hybrid_alltoall_any 4261 16.3 0.106 0.376 2.189 2.759 cp_fm_cholesky_invert 11 10.9 2.681 2.687 2.681 2.687 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.613 2.677 calculate_dm_sparse 110 9.5 0.001 0.001 2.616 2.646 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.920 2.530 potential_pw2rs 110 12.3 0.010 0.010 2.457 2.494 mp_waitany 17072 13.8 1.255 2.484 1.255 2.484 grid_collocate_task_list 110 9.6 2.093 2.473 2.093 2.473 mp_irecv_dv 48980 15.7 0.850 2.404 0.850 2.404 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.328 2.403 mp_alltoall_d11v 2046 13.8 1.776 2.278 1.776 2.278 mp_sum_l 6514 12.8 1.589 2.256 1.589 2.256 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.179 0.926 2.120 qs_energies_init_hamiltonians 11 5.9 0.013 0.025 2.036 2.038 dbcsr_complete_redistribute 325 12.2 0.337 0.382 1.425 1.890 cp_fm_upper_to_full 70 13.6 1.393 1.859 1.393 1.859 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.748 1.764 jit_kernel_multiply 8 16.8 0.538 1.607 0.538 1.607 cp_fm_cholesky_decompose 22 10.9 1.585 1.605 1.585 1.605 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.570 1.584 mp_allgather_i34 2055 14.4 0.477 1.582 0.477 1.582 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.371 1.498 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.973 1.426 rs_gather_matrices 110 12.3 0.234 0.260 0.966 1.374 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.263 1.270 acc_transpose_blocks 16440 15.4 0.072 0.073 1.259 1.267 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.533000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.454545, yerr=8.783907 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 741.543936E+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_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 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 0.057 0.130 65.895 65.916 qs_mol_dyn_low 1 2.0 0.003 0.004 65.497 65.507 qs_forces 11 3.9 0.006 0.030 65.307 65.309 qs_energies 11 4.9 0.002 0.005 61.863 61.867 scf_env_do_scf 11 5.9 0.002 0.016 53.169 53.172 scf_env_do_scf_inner_loop 99 6.5 0.005 0.023 41.488 41.491 velocity_verlet 10 3.0 0.001 0.001 37.109 37.116 dbcsr_multiply_generic 2055 12.4 0.119 0.150 29.521 29.708 qs_scf_new_mos 99 7.5 0.001 0.001 26.251 26.354 qs_scf_loop_do_ot 99 8.5 0.001 0.002 26.250 26.353 ot_scf_mini 99 9.5 0.003 0.004 24.629 24.719 multiply_cannon 2055 13.4 0.259 0.287 22.374 23.458 multiply_cannon_loop 2055 14.4 0.889 0.909 21.006 21.645 ot_mini 99 10.5 0.001 0.001 13.987 14.106 multiply_cannon_multrec 24660 15.4 4.242 6.965 12.689 13.823 rebuild_ks_matrix 110 8.3 0.000 0.000 12.286 12.382 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.285 12.382 init_scf_loop 11 6.9 0.003 0.026 11.638 11.638 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.859 10.946 qs_ot_get_derivative 99 11.5 0.001 0.001 9.824 9.921 prepare_preconditioner 11 7.9 0.000 0.000 9.879 9.895 make_preconditioner 11 8.9 0.000 0.001 9.879 9.895 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.120 9.551 dbcsr_mm_accdrv_process 52304 16.0 6.905 8.726 8.300 9.272 sum_up_and_integrate 110 10.3 0.068 0.071 6.592 6.607 integrate_v_rspace 110 11.3 0.003 0.003 6.524 6.538 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.238 6.248 calculate_rho_elec 110 8.6 0.077 0.081 6.238 6.247 mp_waitall_1 121746 16.5 4.219 6.183 4.219 6.183 qs_ot_get_p 110 10.4 0.001 0.001 5.697 5.847 init_scf_run 11 5.9 0.000 0.001 5.839 5.839 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.838 5.839 make_m2s 4110 13.4 0.065 0.111 5.492 5.755 make_images 4110 14.4 0.583 0.747 5.346 5.614 cp_fm_upper_to_full 70 13.8 3.251 4.612 3.251 4.612 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.048 4.145 apply_single 110 13.6 0.000 0.000 4.048 4.145 ot_diis_step 99 11.5 0.011 0.011 4.127 4.127 pw_transfer 1331 11.6 0.065 0.074 3.947 3.975 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.922 3.936 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.841 3.872 dbcsr_complete_redistribute 325 12.2 0.434 0.492 2.682 3.810 density_rs2pw 110 9.6 0.004 0.005 3.136 3.647 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.449 3.450 grid_integrate_task_list 110 12.3 3.258 3.432 3.258 3.432 fft_wrap_pw1pw2_140 451 13.1 0.202 0.212 3.298 3.334 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.255 3.308 multiply_cannon_sync_h2d 24660 15.4 3.161 3.288 3.161 3.288 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.148 3.254 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.383 3.243 fft3d_ps 1111 14.6 1.091 1.125 3.131 3.156 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.148 3.150 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.004 3.057 wfi_extrapolate 11 7.9 0.001 0.001 2.977 2.977 make_images_data 4110 15.4 0.047 0.051 2.654 2.958 rs_pw_transfer 902 11.9 0.010 0.011 2.435 2.956 calculate_dm_sparse 110 9.5 0.001 0.001 2.910 2.936 hybrid_alltoall_any 4261 16.3 0.120 0.460 2.357 2.921 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.749 2.847 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.839 2.842 cp_fm_redistribute_end 48 14.0 1.420 2.812 1.422 2.813 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.757 2.797 cp_fm_diag_elpa_base 48 14.0 1.310 2.677 1.388 2.788 calculate_first_density_matrix 1 7.0 0.000 0.000 2.766 2.770 mp_alltoall_i22 605 13.7 1.631 2.763 1.631 2.763 cp_fm_cholesky_invert 11 10.9 2.728 2.738 2.728 2.738 grid_collocate_task_list 110 9.6 2.184 2.437 2.184 2.437 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.284 2.285 potential_pw2rs 110 12.3 0.013 0.013 2.244 2.254 mp_alltoall_d11v 2046 13.8 1.835 2.228 1.835 2.228 jit_kernel_multiply 12 15.9 1.062 2.201 1.062 2.201 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.699 1.737 cp_fm_cholesky_decompose 22 10.9 1.659 1.710 1.659 1.710 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.600 1.700 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.673 1.683 mp_sum_l 6514 12.8 1.013 1.624 1.013 1.624 mp_waitany 13376 13.8 1.080 1.589 1.080 1.589 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.523 1.539 acc_transpose_blocks 24660 15.4 0.108 0.110 1.502 1.523 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.841 1.484 mp_allgather_i34 2055 14.4 0.446 1.477 0.446 1.477 mp_irecv_dv 62702 16.1 0.742 1.409 0.742 1.409 rs_pw_transfer_RS2PW_140 121 11.5 0.168 0.174 0.830 1.353 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.916000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=699.818182, yerr=14.211507 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 834.183168E+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_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 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 0.012 0.027 56.186 56.186 qs_mol_dyn_low 1 2.0 0.003 0.003 55.961 55.971 qs_forces 11 3.9 0.002 0.002 55.897 55.898 qs_energies 11 4.9 0.001 0.001 52.183 52.188 scf_env_do_scf 11 5.9 0.000 0.001 43.894 43.894 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.972 35.973 velocity_verlet 10 3.0 0.001 0.001 31.518 31.521 dbcsr_multiply_generic 2055 12.4 0.106 0.108 22.841 23.004 qs_scf_new_mos 99 7.5 0.001 0.001 20.626 20.667 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.625 20.666 ot_scf_mini 99 9.5 0.002 0.002 19.387 19.418 multiply_cannon 2055 13.4 0.243 0.257 17.395 18.668 multiply_cannon_loop 2055 14.4 0.322 0.333 16.043 16.378 rebuild_ks_matrix 110 8.3 0.000 0.000 12.144 12.193 qs_ks_build_kohn_sham_matrix 110 9.3 0.040 0.059 12.143 12.193 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.793 10.838 ot_mini 99 10.5 0.001 0.001 10.485 10.514 multiply_cannon_multrec 8220 15.4 3.206 4.724 7.421 8.729 mp_waitall_1 103326 16.6 6.164 7.882 6.164 7.882 init_scf_loop 11 6.9 0.000 0.000 7.874 7.875 qs_ot_get_derivative 99 11.5 0.001 0.001 6.631 6.663 sum_up_and_integrate 110 10.3 0.079 0.081 6.630 6.643 integrate_v_rspace 110 11.3 0.003 0.003 6.550 6.563 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.435 6.449 calculate_rho_elec 110 8.6 0.115 0.116 6.434 6.449 prepare_preconditioner 11 7.9 0.000 0.000 6.169 6.196 make_preconditioner 11 8.9 0.000 0.000 6.169 6.196 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.755 5.827 init_scf_run 11 5.9 0.000 0.001 5.131 5.131 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.131 5.131 qs_ot_get_p 110 10.4 0.001 0.001 4.985 5.020 dbcsr_mm_accdrv_process 17442 15.9 2.764 3.447 4.085 5.006 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.175 4.674 make_m2s 4110 13.4 0.038 0.040 4.161 4.400 make_images 4110 14.4 0.639 0.696 4.031 4.270 pw_transfer 1331 11.6 0.066 0.072 4.210 4.223 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.102 4.120 ot_diis_step 99 11.5 0.012 0.012 3.832 3.833 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.753 3.801 apply_single 110 13.6 0.000 0.000 3.753 3.801 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.638 3.642 grid_integrate_task_list 110 12.3 3.346 3.570 3.346 3.570 fft_wrap_pw1pw2_140 451 13.1 0.216 0.218 3.545 3.563 density_rs2pw 110 9.6 0.004 0.004 3.174 3.552 fft3d_ps 1111 14.6 1.148 1.171 3.345 3.359 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.333 3.334 multiply_cannon_sync_h2d 8220 15.4 2.902 3.056 2.902 3.056 cp_fm_cholesky_invert 11 10.9 2.999 3.002 2.999 3.002 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.815 2.816 cp_fm_redistribute_end 48 14.0 0.718 2.789 0.722 2.789 wfi_extrapolate 11 7.9 0.001 0.001 2.770 2.770 make_images_data 4110 15.4 0.038 0.043 2.370 2.762 cp_fm_diag_elpa_base 48 14.0 1.883 2.589 2.060 2.755 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.708 2.708 hybrid_alltoall_any 4261 16.3 0.200 0.864 2.323 2.674 rs_pw_transfer 902 11.9 0.010 0.010 2.202 2.608 grid_collocate_task_list 110 9.6 2.278 2.538 2.278 2.538 calculate_dm_sparse 110 9.5 0.001 0.001 2.462 2.500 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.487 2.489 calculate_first_density_matrix 1 7.0 0.000 0.000 2.262 2.264 potential_pw2rs 110 12.3 0.015 0.016 2.161 2.168 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.144 2.162 mp_alltoall_d11v 2046 13.8 1.690 2.121 1.690 2.121 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.992 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.891 1.910 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.810 1.822 cp_fm_cholesky_decompose 22 10.9 1.683 1.705 1.683 1.705 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.688 1.692 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.543 1.669 mp_allgather_i34 2055 14.4 0.540 1.620 0.540 1.620 jit_kernel_multiply 9 15.9 1.007 1.551 1.007 1.551 dbcsr_complete_redistribute 325 12.2 0.555 0.583 1.434 1.524 mp_waitany 9240 13.8 1.060 1.485 1.060 1.485 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.425 1.436 rs_gather_matrices 110 12.3 0.326 0.364 0.953 1.361 qs_create_task_list 11 7.9 0.004 0.031 1.252 1.357 generate_qs_task_list 11 8.9 0.379 0.446 1.247 1.357 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.482 1.337 mp_irecv_dv 24056 15.7 0.457 1.297 0.457 1.297 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.170 1.208 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.167 0.794 1.193 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.186000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=782.545455, yerr=11.324660 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 1.366065E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.061 0.106 88.643 88.647 qs_mol_dyn_low 1 2.0 0.003 0.003 88.249 88.258 qs_forces 11 3.9 0.002 0.002 88.181 88.182 qs_energies 11 4.9 0.007 0.008 83.974 83.976 scf_env_do_scf 11 5.9 0.001 0.001 73.809 73.809 velocity_verlet 10 3.0 0.001 0.001 56.422 56.428 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 45.166 45.168 dbcsr_multiply_generic 2055 12.4 0.125 0.127 28.703 28.734 init_scf_loop 11 6.9 0.000 0.000 28.569 28.572 qs_scf_new_mos 99 7.5 0.001 0.001 26.484 26.520 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.483 26.519 prepare_preconditioner 11 7.9 0.000 0.000 26.468 26.475 make_preconditioner 11 8.9 0.000 0.000 26.468 26.475 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.547 25.945 ot_scf_mini 99 9.5 0.002 0.002 24.716 24.745 multiply_cannon 2055 13.4 0.346 0.375 21.492 22.222 multiply_cannon_loop 2055 14.4 0.341 0.342 19.689 20.096 cp_fm_upper_to_full 70 14.2 12.647 18.296 12.647 18.296 rebuild_ks_matrix 110 8.3 0.001 0.001 14.465 14.507 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.464 14.507 ot_mini 99 10.5 0.001 0.001 13.611 13.637 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.113 13.153 dbcsr_complete_redistribute 325 12.2 1.034 1.055 7.544 10.757 multiply_cannon_multrec 8220 15.4 4.374 4.589 9.603 9.712 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.491 9.705 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.905 9.099 qs_ot_get_derivative 99 11.5 0.001 0.001 9.010 9.041 mp_alltoall_i22 605 13.7 5.527 8.738 5.527 8.738 mp_waitall_1 84994 16.7 7.685 8.496 7.685 8.496 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.903 7.941 calculate_rho_elec 110 8.6 0.227 0.227 7.902 7.940 sum_up_and_integrate 110 10.3 0.150 0.152 7.516 7.534 integrate_v_rspace 110 11.3 0.003 0.004 7.366 7.382 make_m2s 4110 13.4 0.043 0.043 5.452 5.969 qs_ot_get_p 110 10.4 0.001 0.001 5.877 5.897 init_scf_run 11 5.9 0.000 0.001 5.849 5.849 scf_env_initial_rho_setup 11 6.9 0.012 0.013 5.849 5.849 make_images 4110 14.4 0.878 0.919 5.264 5.780 pw_transfer 1331 11.6 0.075 0.075 5.418 5.420 cp_fm_cholesky_invert 11 10.9 5.345 5.350 5.345 5.350 dbcsr_mm_accdrv_process 11614 15.7 3.222 3.597 5.087 5.339 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.301 5.303 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.725 5.180 apply_single 110 13.6 0.000 0.000 4.725 5.180 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.735 5.143 fft_wrap_pw1pw2_140 451 13.1 0.222 0.223 4.601 4.605 ot_diis_step 99 11.5 0.015 0.015 4.578 4.578 fft3d_ps 1111 14.6 1.296 1.307 4.493 4.497 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.301 4.308 multiply_cannon_sync_h2d 8220 15.4 3.948 3.955 3.948 3.955 density_rs2pw 110 9.6 0.004 0.004 3.916 3.937 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.855 3.856 hybrid_alltoall_any 4261 16.3 0.258 0.560 3.032 3.809 make_images_data 4110 15.4 0.042 0.045 3.043 3.717 grid_integrate_task_list 110 12.3 3.684 3.716 3.684 3.716 qs_energies_init_hamiltonians 11 5.9 0.013 0.024 3.703 3.704 wfi_extrapolate 11 7.9 0.001 0.001 3.330 3.330 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.858 3.314 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.242 3.242 cp_fm_diag_elpa_base 48 14.0 2.701 2.894 3.240 3.241 calculate_dm_sparse 110 9.5 0.001 0.001 3.129 3.147 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.975 2.979 grid_collocate_task_list 110 9.6 2.651 2.690 2.651 2.690 potential_pw2rs 110 12.3 0.021 0.022 2.575 2.582 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.471 2.484 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.399 2.413 calculate_first_density_matrix 1 7.0 0.000 0.000 2.384 2.385 rs_pw_transfer 902 11.9 0.011 0.011 2.331 2.364 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.210 2.262 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.089 2.175 mp_alltoall_d11v 2046 13.8 1.985 2.106 1.985 2.106 cp_fm_cholesky_decompose 22 10.9 2.013 2.032 2.013 2.032 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.958 1.961 qs_create_task_list 11 7.9 0.007 0.014 1.908 1.955 generate_qs_task_list 11 8.9 0.734 0.787 1.901 1.941 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.780 1.809 jit_kernel_multiply 10 15.4 1.661 1.800 1.661 1.800 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.647000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1214.363636, yerr=65.637119 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 627.830784E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2533274. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 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 0.012 0.029 204.644 204.645 qs_mol_dyn_low 1 2.0 0.003 0.003 204.284 204.297 qs_forces 11 3.9 0.003 0.003 204.197 204.198 qs_energies 11 4.9 0.001 0.002 198.601 198.625 scf_env_do_scf 11 5.9 0.001 0.001 181.939 181.944 scf_env_do_scf_inner_loop 117 6.6 0.002 0.008 161.148 161.150 dbcsr_multiply_generic 2507 12.6 0.176 0.182 123.436 123.758 velocity_verlet 10 3.0 0.001 0.001 123.310 123.311 qs_scf_new_mos 117 7.6 0.001 0.001 120.737 120.940 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.736 120.940 ot_scf_mini 117 9.6 0.003 0.003 114.171 114.356 multiply_cannon 2507 13.6 0.235 0.243 100.359 101.935 multiply_cannon_loop 2507 14.6 2.123 2.182 98.179 99.944 ot_mini 117 10.6 0.001 0.001 65.255 65.459 multiply_cannon_multrec 60168 15.6 32.983 34.640 41.366 43.010 qs_ot_get_derivative 117 11.6 0.001 0.001 40.561 40.744 rebuild_ks_matrix 128 8.3 0.001 0.001 33.941 34.195 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.940 34.194 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.530 30.788 mp_waitall_1 267128 16.5 28.076 30.512 28.076 30.512 multiply_cannon_sync_h2d 60168 15.6 27.272 28.471 27.272 28.471 qs_ot_get_p 128 10.4 0.001 0.001 27.077 27.291 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.224 24.771 apply_single 128 13.6 0.001 0.001 24.224 24.770 ot_diis_step 117 11.6 0.007 0.008 24.295 24.296 init_scf_loop 11 6.9 0.000 0.000 20.714 20.715 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.487 20.577 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.688 18.886 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.945 17.946 multiply_cannon_metrocomm3 60168 15.6 0.119 0.123 15.424 17.738 prepare_preconditioner 11 7.9 0.000 0.000 16.065 16.119 make_preconditioner 11 8.9 0.000 0.000 16.065 16.119 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.271 15.455 cp_fm_diag_elpa 83 13.4 0.000 0.000 15.076 15.082 cp_fm_redistribute_end 83 14.4 11.914 14.997 11.927 15.003 cp_fm_diag_elpa_base 83 14.4 3.028 14.616 3.061 14.734 sum_up_and_integrate 128 10.3 0.089 0.107 14.446 14.463 integrate_v_rspace 128 11.3 0.003 0.004 14.356 14.379 make_m2s 5014 13.6 0.104 0.112 13.869 14.260 make_images 5014 14.6 0.401 0.420 13.687 14.084 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.909 14.028 calculate_rho_elec 128 8.7 0.045 0.064 13.909 14.027 init_scf_run 11 5.9 0.000 0.001 12.500 12.500 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.499 12.500 density_rs2pw 128 9.7 0.006 0.007 7.309 10.544 cp_fm_cholesky_invert 11 10.9 9.218 9.226 9.218 9.226 wfi_extrapolate 11 7.9 0.001 0.001 9.195 9.195 rs_pw_transfer 1046 11.9 0.016 0.018 5.900 9.151 mp_sum_l 7870 13.0 7.941 9.024 7.941 9.024 calculate_dm_sparse 128 9.5 0.001 0.001 8.364 8.526 dbcsr_mm_accdrv_process 124484 16.2 3.321 3.462 7.947 8.470 pw_transfer 1547 11.6 0.077 0.101 8.075 8.320 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 7.869 8.118 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.955 8.089 multiply_cannon_metrocomm1 60168 15.6 0.090 0.093 6.047 8.084 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.815 7.929 make_images_data 5014 15.6 0.068 0.074 6.798 7.702 grid_integrate_task_list 128 12.3 7.103 7.567 7.103 7.567 hybrid_alltoall_any 5200 16.5 0.291 2.257 5.954 7.472 fft_wrap_pw1pw2_140 523 13.2 0.443 0.510 6.779 7.079 fft3d_ps 1291 14.7 2.105 2.879 6.658 6.832 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.634 6.641 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.804 5.909 mp_waitany 16020 13.9 2.676 5.908 2.676 5.908 grid_collocate_task_list 128 9.7 4.554 5.789 4.554 5.789 rs_pw_transfer_RS2PW_140 139 11.5 0.278 0.297 2.105 5.376 mp_alltoall_d11v 2415 14.1 4.149 5.308 4.149 5.308 potential_pw2rs 128 12.3 0.009 0.011 4.833 4.856 cp_fm_cholesky_decompose 22 10.9 4.634 4.646 4.634 4.646 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.645000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.636364, yerr=6.414086 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 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 11444707676160 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 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 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 828.440576E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453843137296 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4083 57228. MP_Allreduce 11247 947. MP_Sync 168 MP_Alltoall 1955 5620301. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 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 0.020 0.039 191.846 191.849 qs_mol_dyn_low 1 2.0 0.003 0.005 191.392 191.405 qs_forces 11 3.9 0.003 0.005 190.961 190.963 qs_energies 11 4.9 0.002 0.007 184.140 184.152 scf_env_do_scf 11 5.9 0.001 0.001 167.636 167.646 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 134.421 134.424 velocity_verlet 10 3.0 0.001 0.001 120.151 120.161 dbcsr_multiply_generic 2485 12.5 0.182 0.188 96.339 97.595 qs_scf_new_mos 116 7.6 0.001 0.001 94.323 94.974 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.323 94.973 ot_scf_mini 116 9.6 0.004 0.004 89.555 90.234 multiply_cannon 2485 13.5 0.474 0.526 76.589 80.841 multiply_cannon_loop 2485 14.5 1.243 1.297 73.607 76.376 ot_mini 116 10.6 0.001 0.001 49.500 50.126 mp_waitall_1 212858 16.6 24.328 37.363 24.328 37.363 multiply_cannon_multrec 29820 15.5 21.774 25.782 31.388 35.672 rebuild_ks_matrix 127 8.3 0.001 0.001 33.224 33.628 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 33.224 33.627 init_scf_loop 11 6.9 0.006 0.028 33.127 33.129 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.888 30.256 prepare_preconditioner 11 7.9 0.000 0.000 28.661 28.744 make_preconditioner 11 8.9 0.000 0.000 28.661 28.744 multiply_cannon_metrocomm3 29820 15.5 0.094 0.098 15.674 28.503 qs_ot_get_derivative 116 11.6 0.001 0.001 27.699 28.370 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.339 27.872 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.887 22.957 apply_single 127 13.6 0.001 0.001 21.886 22.957 qs_ot_get_p 127 10.4 0.001 0.001 22.056 22.892 ot_diis_step 116 11.6 0.014 0.015 21.636 21.638 multiply_cannon_sync_h2d 29820 15.5 19.107 21.366 19.107 21.366 qs_ot_p2m_diag 82 11.4 0.185 0.212 17.333 17.372 cp_fm_cholesky_invert 11 10.9 16.855 16.868 16.855 16.868 cp_dbcsr_syevd 82 12.4 0.005 0.006 16.184 16.185 make_m2s 4970 13.5 0.088 0.093 14.124 15.382 make_images 4970 14.5 1.156 1.353 13.915 15.174 sum_up_and_integrate 127 10.3 0.115 0.130 14.804 14.838 integrate_v_rspace 127 11.3 0.003 0.003 14.689 14.724 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.242 14.275 calculate_rho_elec 127 8.7 0.087 0.105 14.241 14.275 cp_fm_diag_elpa 82 13.4 0.000 0.001 13.023 13.033 cp_fm_redistribute_end 82 14.4 7.602 12.956 7.615 12.961 cp_fm_diag_elpa_base 82 14.4 5.110 12.453 5.325 12.806 init_scf_run 11 5.9 0.000 0.001 11.566 11.568 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.566 11.567 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.868 11.388 multiply_cannon_metrocomm4 27335 15.5 0.093 0.107 3.766 10.516 mp_irecv_dv 68888 16.3 3.575 10.131 3.575 10.131 density_rs2pw 127 9.7 0.006 0.007 7.572 10.119 make_images_data 4970 15.5 0.064 0.072 8.548 10.060 dbcsr_mm_accdrv_process 61726 16.2 4.571 5.414 9.076 9.699 hybrid_alltoall_any 5155 16.4 0.340 1.505 7.450 9.398 pw_transfer 1535 11.6 0.087 0.111 9.047 9.129 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.822 8.909 wfi_extrapolate 11 7.9 0.001 0.001 8.396 8.396 rs_pw_transfer 1038 11.9 0.014 0.016 5.759 8.318 fft_wrap_pw1pw2_140 519 13.2 0.469 0.518 7.771 7.861 grid_integrate_task_list 127 12.3 7.094 7.732 7.094 7.732 fft3d_ps 1281 14.7 2.734 2.899 7.199 7.247 cp_fm_cholesky_decompose 22 10.9 6.919 7.002 6.919 7.002 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.177 6.880 calculate_dm_sparse 127 9.5 0.001 0.001 6.489 6.618 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.098 6.111 mp_sum_l 7804 13.0 3.867 5.900 3.867 5.900 grid_collocate_task_list 127 9.7 4.666 5.822 4.666 5.822 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.328 5.396 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.219 5.380 mp_alltoall_d11v 2401 14.1 4.453 5.216 4.453 5.216 potential_pw2rs 127 12.3 0.014 0.016 5.109 5.133 mp_waitany 11660 13.9 2.485 5.083 2.485 5.083 rs_pw_transfer_RS2PW_140 138 11.5 0.347 0.375 2.083 4.639 mp_allgather_i34 2485 14.5 1.562 4.501 1.562 4.501 mp_sum_d 4441 12.1 2.705 4.097 2.705 4.097 dbcsr_complete_redistribute 393 12.7 0.785 0.888 3.159 3.984 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.849000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.727273, yerr=1.285649 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 929.955840E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.030 179.500 179.501 qs_mol_dyn_low 1 2.0 0.003 0.004 178.997 179.122 qs_forces 11 3.9 0.003 0.003 178.893 178.898 qs_energies 11 4.9 0.001 0.002 172.277 172.286 scf_env_do_scf 11 5.9 0.001 0.001 156.702 156.703 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 121.306 121.307 velocity_verlet 10 3.0 0.001 0.001 114.319 114.321 dbcsr_multiply_generic 2507 12.6 0.183 0.189 82.296 83.342 qs_scf_new_mos 117 7.6 0.001 0.001 82.501 82.858 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.501 82.857 ot_scf_mini 117 9.6 0.003 0.004 78.357 78.755 multiply_cannon 2507 13.6 0.501 0.520 62.417 66.940 multiply_cannon_loop 2507 14.6 0.855 0.882 59.378 61.897 ot_mini 117 10.6 0.001 0.001 43.236 43.608 init_scf_loop 11 6.9 0.000 0.000 35.295 35.297 mp_waitall_1 170520 16.6 25.248 34.729 25.248 34.729 rebuild_ks_matrix 128 8.3 0.001 0.001 31.131 31.549 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 31.130 31.548 prepare_preconditioner 11 7.9 0.000 0.000 31.166 31.221 make_preconditioner 11 8.9 0.000 0.000 31.166 31.221 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.811 30.237 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.071 28.453 multiply_cannon_multrec 20056 15.6 13.415 16.482 22.849 26.021 multiply_cannon_metrocomm3 20056 15.6 0.061 0.065 15.302 24.377 qs_ot_get_derivative 117 11.6 0.001 0.002 23.394 23.780 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.973 21.031 apply_single 128 13.6 0.001 0.001 19.972 21.031 qs_ot_get_p 128 10.4 0.001 0.001 20.505 20.933 ot_diis_step 117 11.6 0.018 0.018 19.735 19.736 qs_ot_p2m_diag 83 11.4 0.265 0.272 16.214 16.227 make_m2s 5014 13.6 0.080 0.085 14.848 15.873 multiply_cannon_sync_h2d 20056 15.6 14.242 15.838 14.242 15.838 make_images 5014 14.6 1.180 1.256 14.614 15.639 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.187 15.188 sum_up_and_integrate 128 10.3 0.134 0.145 14.903 14.931 integrate_v_rspace 128 11.3 0.003 0.004 14.769 14.797 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.626 14.654 calculate_rho_elec 128 8.7 0.132 0.147 14.626 14.653 cp_fm_cholesky_invert 11 10.9 14.480 14.489 14.480 14.489 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.935 11.936 cp_fm_redistribute_end 83 14.4 4.511 11.863 4.525 11.866 cp_fm_diag_elpa_base 83 14.4 6.905 11.270 7.317 11.734 make_images_data 5014 15.6 0.061 0.068 9.216 10.696 init_scf_run 11 5.9 0.000 0.001 10.574 10.574 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.574 10.574 hybrid_alltoall_any 5200 16.5 0.433 1.990 7.846 9.660 density_rs2pw 128 9.7 0.006 0.006 7.474 9.465 pw_transfer 1547 11.6 0.087 0.109 9.235 9.352 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.935 9.211 multiply_cannon_metrocomm4 17549 15.6 0.061 0.071 3.478 9.137 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 9.009 9.135 dbcsr_mm_accdrv_process 41502 16.2 4.705 5.978 8.892 9.065 mp_irecv_dv 50230 16.2 3.356 8.878 3.356 8.878 fft_wrap_pw1pw2_140 523 13.2 0.476 0.521 7.922 8.058 grid_integrate_task_list 128 12.3 7.279 7.941 7.279 7.941 cp_fm_upper_to_full 105 14.5 5.834 7.552 5.834 7.552 cp_fm_cholesky_decompose 22 10.9 7.453 7.483 7.453 7.483 wfi_extrapolate 11 7.9 0.001 0.001 7.435 7.435 rs_pw_transfer 1046 11.9 0.014 0.014 5.395 7.409 fft3d_ps 1291 14.7 2.666 2.901 7.274 7.356 dbcsr_complete_redistribute 395 12.7 1.170 1.196 4.607 6.410 calculate_dm_sparse 128 9.5 0.001 0.001 5.852 5.952 grid_collocate_task_list 128 9.7 4.906 5.819 4.906 5.819 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.489 5.496 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.562 5.223 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.412 5.211 mp_alltoall_d11v 2415 14.1 4.378 5.014 4.378 5.014 potential_pw2rs 128 12.3 0.020 0.023 4.991 5.006 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.592 4.714 mp_sum_l 7870 13.0 3.208 4.658 3.208 4.658 mp_allgather_i34 2507 14.6 1.561 4.572 1.561 4.572 mp_waitany 11748 13.9 2.384 4.504 2.384 4.504 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.335 4.116 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.059 4.105 rs_pw_transfer_RS2PW_140 139 11.5 0.332 0.362 1.948 3.956 mp_alltoall_i22 716 14.1 1.936 3.880 1.936 3.880 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 3.768 3.770 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.574 3.615 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=179.501000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=882.000000, yerr=9.184968 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.158300E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56626. MP_Allreduce 11356 1063. MP_Sync 172 MP_Alltoall 1724 12509575. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_comm_split 84 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 0.046 0.088 191.399 191.403 qs_mol_dyn_low 1 2.0 0.003 0.005 190.952 190.966 qs_forces 11 3.9 0.003 0.011 190.849 190.854 qs_energies 11 4.9 0.002 0.008 183.629 183.641 scf_env_do_scf 11 5.9 0.002 0.014 166.715 166.724 velocity_verlet 10 3.0 0.001 0.001 126.521 126.524 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 119.877 119.878 qs_scf_new_mos 118 7.6 0.001 0.001 81.539 81.865 qs_scf_loop_do_ot 118 8.6 0.001 0.002 81.538 81.865 dbcsr_multiply_generic 2529 12.6 0.191 0.196 80.833 81.667 ot_scf_mini 118 9.6 0.003 0.004 77.026 77.316 multiply_cannon 2529 13.6 0.561 0.584 55.926 58.694 multiply_cannon_loop 2529 14.6 1.193 1.221 52.212 54.011 init_scf_loop 11 6.9 0.001 0.005 46.704 46.705 ot_mini 118 10.6 0.001 0.001 43.492 43.780 prepare_preconditioner 11 7.9 0.000 0.000 42.437 42.480 make_preconditioner 11 8.9 0.000 0.000 42.436 42.480 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.905 40.916 multiply_cannon_multrec 30348 15.6 14.180 18.995 26.471 31.680 rebuild_ks_matrix 129 8.3 0.001 0.001 30.483 30.729 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 30.483 30.728 mp_waitall_1 149172 16.7 18.154 28.662 18.154 28.662 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.500 27.722 qs_ot_get_derivative 118 11.6 0.001 0.002 23.553 23.833 make_m2s 5058 13.6 0.095 0.101 20.490 21.488 make_images 5058 14.6 1.971 2.285 20.182 21.178 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.371 20.045 apply_single 129 13.6 0.001 0.001 19.371 20.045 qs_ot_get_p 129 10.4 0.001 0.001 19.631 19.967 ot_diis_step 118 11.6 0.018 0.019 19.813 19.815 cp_fm_upper_to_full 106 14.7 11.001 16.324 11.001 16.324 cp_fm_cholesky_invert 11 10.9 16.241 16.249 16.241 16.249 multiply_cannon_metrocomm3 30348 15.6 0.047 0.050 7.042 15.837 qs_ot_p2m_diag 84 11.4 0.346 0.394 15.568 15.620 sum_up_and_integrate 129 10.3 0.143 0.155 15.040 15.062 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.927 14.960 calculate_rho_elec 129 8.7 0.176 0.192 14.927 14.959 integrate_v_rspace 129 11.3 0.004 0.004 14.897 14.925 cp_dbcsr_syevd 84 12.4 0.005 0.006 14.267 14.268 dbcsr_complete_redistribute 397 12.7 1.508 1.625 9.142 12.958 make_images_data 5058 15.6 0.065 0.070 11.082 12.791 dbcsr_mm_accdrv_process 62780 16.2 7.469 9.144 11.860 12.668 multiply_cannon_sync_h2d 30348 15.6 11.845 12.589 11.845 12.589 copy_fm_to_dbcsr 210 11.7 0.002 0.002 7.763 11.571 hybrid_alltoall_any 5245 16.5 0.532 2.227 9.881 11.523 cp_fm_diag_elpa 84 13.4 0.000 0.001 11.122 11.125 cp_fm_redistribute_end 84 14.4 1.923 11.044 1.938 11.050 init_scf_run 11 5.9 0.000 0.001 10.926 10.927 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.925 10.927 cp_fm_diag_elpa_base 84 14.4 8.492 10.419 9.077 10.910 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.508 10.206 pw_transfer 1559 11.6 0.089 0.108 9.669 9.743 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.437 9.647 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.440 9.518 mp_alltoall_i22 720 14.1 5.610 9.466 5.610 9.466 density_rs2pw 129 9.7 0.006 0.006 7.292 8.543 fft_wrap_pw1pw2_140 527 13.2 0.486 0.497 8.368 8.451 grid_integrate_task_list 129 12.3 7.569 7.994 7.569 7.994 multiply_cannon_metrocomm4 25290 15.6 0.076 0.088 2.817 7.818 wfi_extrapolate 11 7.9 0.001 0.001 7.754 7.754 fft3d_ps 1301 14.7 2.798 2.871 7.652 7.713 cp_fm_cholesky_decompose 22 10.9 7.596 7.677 7.596 7.677 mp_irecv_dv 76751 16.2 2.669 7.540 2.669 7.540 calculate_dm_sparse 129 9.5 0.001 0.001 6.235 6.312 rs_pw_transfer 1054 12.0 0.013 0.014 4.796 6.108 mp_alltoall_d11v 2429 14.1 5.020 6.067 5.020 6.067 grid_collocate_task_list 129 9.7 5.125 5.862 5.125 5.862 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.433 5.481 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.165 5.277 potential_pw2rs 129 12.3 0.022 0.023 4.772 4.793 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.482 4.484 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.367 4.422 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.232 4.281 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=191.403000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1079.727273, yerr=25.251790 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.531785E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58308. MP_Allreduce 10985 1176. MP_Sync 86 MP_Alltoall 1700 18828162. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.067 0.098 174.969 174.975 qs_mol_dyn_low 1 2.0 0.003 0.004 174.470 174.493 qs_forces 11 3.9 0.012 0.043 174.344 174.354 qs_energies 11 4.9 0.009 0.010 166.788 166.806 scf_env_do_scf 11 5.9 0.001 0.001 149.278 149.283 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 113.453 113.456 velocity_verlet 10 3.0 0.001 0.001 113.435 113.440 qs_scf_new_mos 116 7.6 0.001 0.001 74.653 74.753 qs_scf_loop_do_ot 116 8.6 0.001 0.001 74.652 74.752 dbcsr_multiply_generic 2485 12.5 0.180 0.186 74.059 74.631 ot_scf_mini 116 9.6 0.003 0.004 70.261 70.383 multiply_cannon 2485 13.5 0.579 0.620 54.739 58.686 multiply_cannon_loop 2485 14.5 0.444 0.458 50.100 51.220 ot_mini 116 10.6 0.001 0.001 39.484 39.604 init_scf_loop 11 6.9 0.000 0.000 35.673 35.679 mp_waitall_1 124680 16.7 25.796 33.250 25.796 33.250 prepare_preconditioner 11 7.9 0.000 0.000 31.695 31.740 make_preconditioner 11 8.9 0.000 0.000 31.695 31.740 rebuild_ks_matrix 127 8.3 0.001 0.001 29.953 30.091 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.019 29.953 30.090 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.516 29.781 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.206 27.331 multiply_cannon_multrec 9940 15.5 10.362 15.780 17.967 21.810 qs_ot_get_derivative 116 11.6 0.001 0.002 19.663 19.777 ot_diis_step 116 11.6 0.019 0.020 19.751 19.752 multiply_cannon_metrocomm3 9940 15.5 0.023 0.025 12.705 19.702 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.313 19.668 apply_single 127 13.6 0.001 0.001 19.313 19.668 cp_fm_cholesky_invert 11 10.9 18.110 18.116 18.110 18.116 qs_ot_get_p 127 10.4 0.001 0.001 17.954 18.103 make_m2s 4970 13.5 0.065 0.070 15.705 17.887 make_images 4970 14.5 2.259 2.787 15.402 17.584 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.514 15.554 calculate_rho_elec 127 8.7 0.257 0.267 15.513 15.553 sum_up_and_integrate 127 10.3 0.179 0.188 15.083 15.130 integrate_v_rspace 127 11.3 0.004 0.004 14.904 14.961 qs_ot_p2m_diag 82 11.4 0.489 0.494 14.197 14.214 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.081 13.083 make_images_data 4970 15.5 0.052 0.059 9.524 12.129 multiply_cannon_sync_h2d 9940 15.5 11.436 11.875 11.436 11.875 hybrid_alltoall_any 5155 16.4 0.797 3.472 9.411 11.840 init_scf_run 11 5.9 0.000 0.001 10.645 10.645 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.645 10.645 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.070 10.081 cp_fm_diag_elpa_base 82 14.4 9.830 9.897 10.063 10.074 pw_transfer 1535 11.6 0.085 0.093 9.925 9.959 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.702 9.744 fft_wrap_pw1pw2_140 519 13.2 0.492 0.515 8.570 8.622 grid_integrate_task_list 127 12.3 7.739 8.160 7.739 8.160 cp_fm_cholesky_decompose 22 10.9 7.886 7.967 7.886 7.967 density_rs2pw 127 9.7 0.005 0.006 7.067 7.928 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.831 7.923 fft3d_ps 1281 14.7 2.685 2.772 7.872 7.895 dbcsr_mm_accdrv_process 20590 16.1 2.636 3.492 7.237 7.890 wfi_extrapolate 11 7.9 0.001 0.001 7.385 7.385 multiply_cannon_metrocomm1 9940 15.5 0.029 0.030 4.286 7.296 mp_allgather_i34 2485 14.5 2.857 6.908 2.857 6.908 calculate_dm_sparse 127 9.5 0.001 0.001 6.212 6.289 grid_collocate_task_list 127 9.7 5.353 6.012 5.353 6.012 mp_alltoall_d11v 2401 14.1 5.175 5.825 5.175 5.825 dbcsr_complete_redistribute 393 12.7 2.114 2.167 5.245 5.574 multiply_cannon_metrocomm4 7455 15.5 0.024 0.026 1.869 5.346 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.326 5.336 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.281 5.292 mp_irecv_dv 28618 15.9 1.831 5.257 1.831 5.257 rs_pw_transfer 1038 11.9 0.012 0.013 4.121 5.037 potential_pw2rs 127 12.3 0.026 0.027 4.578 4.595 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.222 4.262 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.562 3.853 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.718 3.754 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.432 3.715 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.650 3.662 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.566 3.628 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.499 3.535 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.975000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1417.090909, yerr=54.875818 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456868864 0.0% 0.0% 100.0% flops 32 x 32 x 32 1960652570624 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 1992007581696 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753954734080 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 11613074706432 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176888320 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176888320 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911117062144 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.230858E+12 0.0% 0.0% 100.0% flops max/rank 11.785799E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806351296 0.0% 0.0% 100.0% number of processed stacks 1979920 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.7 marketing flops 145.648783E+12 ------------------------------------------------------------------------------- # multiplications 2528 max memory usage/rank 3.057750E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101120 MPI messages size (bytes): total size 1.144936E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.322546E+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 45608 35399925760 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592330608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4069 58204. MP_Allreduce 11170 1510. MP_Sync 86 MP_Alltoall 1724 36993574. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.080 300.408 300.412 qs_mol_dyn_low 1 2.0 0.003 0.004 299.776 299.788 qs_forces 11 3.9 0.003 0.003 299.542 299.543 qs_energies 11 4.9 0.002 0.002 290.242 290.251 scf_env_do_scf 11 5.9 0.001 0.001 267.786 267.800 velocity_verlet 10 3.0 0.001 0.002 215.489 215.516 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 143.842 143.844 init_scf_loop 11 6.9 0.000 0.000 123.686 123.689 prepare_preconditioner 11 7.9 0.000 0.000 118.484 118.514 make_preconditioner 11 8.9 0.000 0.000 118.483 118.514 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.247 115.589 qs_scf_new_mos 118 7.6 0.001 0.001 91.993 92.075 qs_scf_loop_do_ot 118 8.6 0.001 0.001 91.992 92.074 ot_scf_mini 118 9.6 0.004 0.004 87.167 87.203 dbcsr_multiply_generic 2528 12.6 0.215 0.221 84.602 85.236 cp_fm_upper_to_full 104 14.8 52.521 74.638 52.521 74.638 multiply_cannon 2528 13.6 0.711 0.739 59.337 59.880 multiply_cannon_loop 2528 14.6 0.479 0.501 55.578 57.150 ot_mini 118 10.6 0.001 0.001 45.274 45.317 dbcsr_complete_redistribute 393 12.7 3.973 4.022 28.874 41.202 rebuild_ks_matrix 129 8.3 0.001 0.001 39.111 39.194 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 39.110 39.193 copy_fm_to_dbcsr 208 11.6 0.001 0.002 25.496 37.767 qs_ks_update_qs_env 129 7.6 0.001 0.001 36.075 36.159 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.193 35.294 cp_fm_cholesky_invert 11 10.9 33.547 33.553 33.547 33.553 mp_alltoall_i22 716 14.1 21.011 33.503 21.011 33.503 mp_waitall_1 104540 16.8 28.963 32.797 28.963 32.797 qs_ot_get_p 129 10.4 0.001 0.001 26.142 26.222 qs_ot_get_derivative 118 11.6 0.002 0.002 24.611 24.644 qs_ot_p2m_diag 82 11.4 0.867 0.872 22.021 22.047 make_m2s 5056 13.6 0.076 0.078 20.571 21.626 make_images 5056 14.6 3.817 3.943 20.089 21.148 qs_rho_update_rho_low 129 7.7 0.001 0.001 21.026 21.039 calculate_rho_elec 129 8.7 0.487 0.487 21.025 21.039 ot_diis_step 118 11.6 0.022 0.023 20.624 20.624 multiply_cannon_metrocomm3 10112 15.6 0.024 0.025 19.022 20.348 cp_dbcsr_syevd 82 12.4 0.005 0.006 20.310 20.314 sum_up_and_integrate 129 10.3 0.323 0.326 20.092 20.161 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.819 20.047 apply_single 129 13.6 0.001 0.001 19.818 20.046 integrate_v_rspace 129 11.3 0.004 0.004 19.768 19.835 multiply_cannon_multrec 10112 15.6 10.597 12.421 18.193 18.265 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.118 17.119 cp_fm_diag_elpa_base 82 14.4 12.814 14.354 17.113 17.113 multiply_cannon_sync_h2d 10112 15.6 15.770 15.788 15.770 15.788 hybrid_alltoall_any 5241 16.5 1.311 3.061 11.291 13.210 make_images_data 5056 15.6 0.061 0.065 11.186 13.157 pw_transfer 1559 11.6 0.094 0.095 12.485 12.493 init_scf_run 11 5.9 0.000 0.001 12.419 12.419 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.418 12.419 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 12.248 12.257 fft_wrap_pw1pw2_140 527 13.2 0.546 0.549 10.871 10.884 fft3d_ps 1301 14.7 2.762 2.771 10.270 10.276 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.441 9.469 dbcsr_mm_accdrv_process 20930 16.1 3.895 5.842 7.355 9.174 mp_alltoall_d11v 2417 14.1 8.351 9.168 8.351 9.168 wfi_extrapolate 11 7.9 0.001 0.001 9.148 9.148 cp_fm_cholesky_decompose 22 10.9 8.901 8.934 8.901 8.934 grid_integrate_task_list 129 12.3 8.635 8.815 8.635 8.815 density_rs2pw 129 9.7 0.005 0.005 8.490 8.636 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.085 8.086 calculate_dm_sparse 129 9.5 0.001 0.001 6.747 6.807 grid_collocate_task_list 129 9.7 6.409 6.445 6.409 6.445 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.331 6.389 rs_scatter_matrices 140 9.7 3.690 4.566 6.094 6.337 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.061 6.135 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=300.412000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2713.000000, yerr=172.184362 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 1.262232E+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 6805335. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.064 85.198 85.199 qs_energies 1 2.0 0.000 0.000 84.669 84.681 ls_scf 1 3.0 0.000 0.000 83.755 83.767 dbcsr_multiply_generic 111 6.7 0.015 0.015 72.738 72.931 multiply_cannon 111 7.7 0.018 0.022 55.947 57.082 multiply_cannon_loop 111 8.7 0.210 0.226 52.550 53.901 ls_scf_main 1 4.0 0.000 0.000 52.449 52.450 density_matrix_trs4 2 5.0 0.002 0.003 46.899 46.961 ls_scf_init_scf 1 4.0 0.000 0.000 28.253 28.254 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.128 27.186 mp_waitall_1 11031 10.9 22.491 25.244 22.491 25.244 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.019 25.037 multiply_cannon_multrec 2664 9.7 8.165 8.941 15.593 17.450 multiply_cannon_sync_h2d 2664 9.7 13.688 15.580 13.688 15.580 make_m2s 222 7.7 0.008 0.010 13.144 13.642 make_images 222 8.7 0.099 0.110 13.122 13.621 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.632 12.397 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.431 8.419 make_images_data 222 9.7 0.004 0.005 7.754 8.273 dbcsr_mm_accdrv_process 4760 10.4 0.516 0.634 7.049 8.079 hybrid_alltoall_any 227 10.6 0.214 1.836 6.625 7.861 dbcsr_mm_accdrv_process_sort 4760 11.4 6.329 7.243 6.329 7.243 calculate_norms 4752 9.8 5.544 6.201 5.544 6.201 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.039 5.189 mp_sum_l 807 5.4 3.130 4.883 3.130 4.883 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.060 3.689 mp_irecv_dv 6231 10.9 2.042 3.664 2.042 3.664 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.307 3.528 make_images_sizes 222 9.7 0.000 0.000 0.737 3.432 mp_alltoall_i44 222 10.7 0.737 3.431 0.737 3.431 arnoldi_extremal 4 6.8 0.000 0.000 3.243 3.272 arnoldi_normal_ev 4 7.8 0.001 0.003 3.242 3.271 build_subspace 16 8.4 0.009 0.012 3.147 3.149 ls_scf_post 1 4.0 0.000 0.000 3.054 3.065 ls_scf_store_result 1 5.0 0.000 0.000 2.843 2.884 dbcsr_special_finalize 555 9.7 0.005 0.006 2.298 2.775 dbcsr_merge_single_wm 555 10.7 0.457 0.596 2.290 2.767 make_images_pack 222 9.7 2.209 2.625 2.211 2.626 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.338 2.580 dbcsr_sort_data 658 11.4 2.090 2.495 2.090 2.495 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.464 2.069 2.466 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.307 2.418 buffer_matrices_ensure_size 222 8.7 1.753 2.070 1.753 2.070 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.796 1.797 rebuild_ks_matrix 3 7.3 0.000 0.000 1.786 1.787 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.786 1.787 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.199000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 2.105033E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.072 91.652 91.652 qs_energies 1 2.0 0.000 0.000 90.976 90.982 ls_scf 1 3.0 0.000 0.000 89.655 89.660 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.681 75.999 multiply_cannon 111 7.7 0.030 0.040 53.288 57.574 ls_scf_main 1 4.0 0.000 0.000 55.530 55.540 multiply_cannon_loop 111 8.7 0.115 0.122 50.016 53.351 density_matrix_trs4 2 5.0 0.002 0.003 49.770 49.979 ls_scf_init_scf 1 4.0 0.000 0.000 30.560 30.562 mp_waitall_1 9105 10.9 21.278 30.061 21.278 30.061 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.342 29.457 multiply_cannon_multrec 1332 9.7 13.220 16.747 22.553 27.192 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.946 26.957 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.695 20.620 make_m2s 222 7.7 0.006 0.007 15.590 16.307 make_images 222 8.7 1.580 1.966 15.560 16.278 dbcsr_mm_accdrv_process 4041 10.4 0.291 0.450 8.927 10.551 dbcsr_mm_accdrv_process_sort 4041 11.4 8.522 10.101 8.522 10.101 make_images_data 222 9.7 0.004 0.004 9.019 9.944 hybrid_alltoall_any 227 10.6 0.521 2.483 8.414 9.525 mp_sum_l 807 5.4 5.504 8.894 5.504 8.894 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.215 7.356 mp_irecv_dv 3311 11.0 3.194 7.303 3.194 7.303 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.286 7.051 calculate_norms 2376 9.8 5.992 6.706 5.992 6.706 multiply_cannon_sync_h2d 1332 9.7 4.800 5.840 4.800 5.840 apply_matrix_preconditioner 6 5.3 0.001 0.002 5.052 5.245 arnoldi_extremal 4 6.8 0.000 0.000 4.605 4.623 arnoldi_normal_ev 4 7.8 0.001 0.005 4.605 4.623 build_subspace 16 8.4 0.014 0.021 4.352 4.355 ls_scf_post 1 4.0 0.000 0.000 3.565 3.572 ls_scf_store_result 1 5.0 0.000 0.000 3.266 3.382 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.125 3.357 dbcsr_matrix_vector_mult_local 304 10.0 2.734 3.215 2.736 3.217 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.647 2.730 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.192 2.693 mp_allgather_i34 111 8.7 0.970 2.504 0.970 2.504 make_images_pack 222 9.7 2.026 2.411 2.028 2.413 dbcsr_sort_data 436 11.2 1.835 2.045 1.835 2.045 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.919 1.920 rebuild_ks_matrix 3 7.3 0.000 0.000 1.906 1.908 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.906 1.907 dbcsr_data_new 4174 10.1 1.610 1.846 1.610 1.846 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.652000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1710.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 2.749637E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.071 0.138 95.288 95.289 qs_energies 1 2.0 0.000 0.000 94.397 94.409 ls_scf 1 3.0 0.002 0.007 92.880 92.891 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.303 77.612 ls_scf_main 1 4.0 0.000 0.000 58.386 58.393 multiply_cannon 111 7.7 0.040 0.100 53.058 57.134 multiply_cannon_loop 111 8.7 0.099 0.106 49.477 53.556 density_matrix_trs4 2 5.0 0.002 0.003 52.404 52.569 mp_waitall_1 7281 11.0 24.283 34.404 24.283 34.404 ls_scf_init_scf 1 4.0 0.003 0.013 30.993 30.995 ls_scf_init_matrix_S 1 5.0 0.001 0.004 29.579 29.651 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.188 27.197 multiply_cannon_multrec 888 9.7 12.571 15.136 21.198 24.273 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.319 23.305 make_m2s 222 7.7 0.006 0.007 17.289 18.567 make_images 222 8.7 1.971 2.278 17.250 18.526 hybrid_alltoall_any 227 10.6 0.620 2.850 9.568 10.985 make_images_data 222 9.7 0.003 0.004 10.000 10.934 dbcsr_mm_accdrv_process 3754 10.4 0.240 0.415 8.163 9.410 dbcsr_mm_accdrv_process_sort 3754 11.4 7.793 8.996 7.793 8.996 mp_sum_l 807 5.4 5.422 8.948 5.422 8.948 multiply_cannon_sync_h2d 888 9.7 6.021 7.336 6.021 7.336 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.473 7.034 mp_irecv_dv 2335 11.1 2.457 6.967 2.457 6.967 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.259 6.961 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.588 6.667 arnoldi_extremal 4 6.8 0.000 0.000 5.105 5.117 arnoldi_normal_ev 4 7.8 0.001 0.005 5.105 5.117 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.865 5.075 build_subspace 16 8.4 0.014 0.020 4.799 4.803 calculate_norms 1584 9.8 4.263 4.581 4.263 4.581 mp_allgather_i34 111 8.7 1.414 3.927 1.414 3.927 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.469 3.786 dbcsr_matrix_vector_mult_local 304 10.0 3.010 3.576 3.012 3.578 ls_scf_post 1 4.0 0.003 0.013 3.499 3.513 ls_scf_store_result 1 5.0 0.000 0.000 3.240 3.323 ls_scf_dm_to_ks 2 5.0 0.000 0.007 2.866 2.972 dbcsr_sort_data 325 11.1 1.907 2.215 1.907 2.215 make_images_pack 222 9.7 1.813 2.139 1.816 2.142 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.138 2.140 rebuild_ks_matrix 3 7.3 0.000 0.000 2.119 2.121 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.015 2.119 2.121 make_images_sizes 222 9.7 0.000 0.000 0.946 2.097 mp_alltoall_i44 222 10.7 0.946 2.097 0.946 2.097 dbcsr_data_release 9322 10.9 1.303 1.926 1.303 1.926 dbcsr_finalize 304 7.8 0.026 0.032 1.612 1.920 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.289000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2198.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 3.329081E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.081 0.118 97.508 97.510 qs_energies 1 2.0 0.000 0.000 96.718 96.724 ls_scf 1 3.0 0.007 0.029 95.029 95.038 dbcsr_multiply_generic 111 6.7 0.017 0.023 78.659 78.915 ls_scf_main 1 4.0 0.000 0.000 59.086 59.097 multiply_cannon 111 7.7 0.049 0.120 51.695 56.073 density_matrix_trs4 2 5.0 0.002 0.003 52.884 53.044 multiply_cannon_loop 111 8.7 0.115 0.126 46.680 50.035 ls_scf_init_scf 1 4.0 0.000 0.001 32.652 32.654 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.369 31.431 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.889 28.904 mp_waitall_1 6369 11.0 22.783 28.645 22.783 28.645 multiply_cannon_multrec 1332 9.7 14.196 17.300 22.157 25.527 make_m2s 222 7.7 0.007 0.009 21.188 22.589 make_images 222 8.7 3.138 3.593 21.138 22.539 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.241 17.267 make_images_data 222 9.7 0.004 0.005 11.836 13.362 hybrid_alltoall_any 227 10.6 0.799 3.810 11.275 12.950 dbcsr_mm_accdrv_process 3641 10.4 0.205 0.408 7.597 9.145 dbcsr_mm_accdrv_process_sort 3641 11.4 7.220 8.729 7.220 8.729 mp_sum_l 807 5.4 4.227 8.373 4.227 8.373 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.167 6.586 multiply_cannon_sync_h2d 1332 9.7 5.546 6.128 5.546 6.128 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.065 6.027 mp_irecv_dv 3229 10.9 2.043 5.952 2.043 5.952 arnoldi_extremal 4 6.8 0.000 0.000 5.232 5.249 arnoldi_normal_ev 4 7.8 0.001 0.004 5.232 5.249 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.538 5.057 build_subspace 16 8.4 0.014 0.021 4.895 4.904 apply_matrix_preconditioner 6 5.3 0.002 0.002 4.603 4.821 calculate_norms 2376 9.8 4.191 4.544 4.191 4.544 mp_allgather_i34 111 8.7 2.139 4.469 2.139 4.469 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.585 3.872 dbcsr_matrix_vector_mult_local 304 10.0 3.177 3.668 3.179 3.669 dbcsr_sort_data 658 11.4 3.144 3.386 3.144 3.386 ls_scf_post 1 4.0 0.001 0.006 3.284 3.292 dbcsr_special_finalize 555 9.7 0.006 0.007 2.888 3.176 dbcsr_merge_single_wm 555 10.7 0.538 0.656 2.879 3.168 ls_scf_dm_to_ks 2 5.0 0.002 0.024 3.044 3.097 ls_scf_store_result 1 5.0 0.000 0.000 2.999 3.074 dbcsr_data_release 10477 10.7 1.587 2.455 1.587 2.455 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.030 2.031 rebuild_ks_matrix 3 7.3 0.000 0.000 2.007 2.008 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.006 2.007 2.008 dbcsr_finalize 304 7.8 0.049 0.062 1.805 1.986 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.510000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2703.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 4.630790E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.078 100.159 100.160 qs_energies 1 2.0 0.000 0.000 99.360 99.367 ls_scf 1 3.0 0.000 0.000 97.412 97.424 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.707 78.952 ls_scf_main 1 4.0 0.000 0.000 62.768 62.769 multiply_cannon 111 7.7 0.076 0.203 55.743 60.236 density_matrix_trs4 2 5.0 0.002 0.003 55.573 55.628 multiply_cannon_loop 111 8.7 0.070 0.078 51.298 53.626 mp_waitall_1 5436 11.0 27.112 32.112 27.112 32.112 ls_scf_init_scf 1 4.0 0.000 0.000 30.965 30.970 ls_scf_init_matrix_S 1 5.0 0.000 0.002 29.634 29.672 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.447 27.455 multiply_cannon_multrec 444 9.7 14.140 16.512 21.269 24.093 make_m2s 222 7.7 0.004 0.005 17.888 20.328 make_images 222 8.7 3.719 4.419 17.826 20.268 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.600 16.613 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.079 15.081 make_images_data 222 9.7 0.003 0.004 10.234 12.494 hybrid_alltoall_any 227 10.6 0.788 3.762 9.985 12.312 multiply_cannon_sync_h2d 444 9.7 6.551 7.983 6.551 7.983 dbcsr_mm_accdrv_process 3003 10.4 0.175 0.338 6.838 7.959 dbcsr_mm_accdrv_process_sort 3003 11.4 6.524 7.617 6.524 7.617 mp_allgather_i34 111 8.7 2.670 7.033 2.670 7.033 arnoldi_extremal 4 6.8 0.000 0.000 5.809 5.826 arnoldi_normal_ev 4 7.8 0.002 0.005 5.808 5.826 build_subspace 16 8.4 0.015 0.020 5.419 5.429 mp_sum_l 807 5.4 3.130 4.748 3.130 4.748 apply_matrix_preconditioner 6 5.3 0.000 0.001 4.585 4.727 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.151 4.361 dbcsr_matrix_vector_mult_local 304 10.0 3.664 4.145 3.666 4.147 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.618 3.894 mp_irecv_dv 1241 11.2 1.600 3.872 1.600 3.872 calculate_norms 792 9.8 3.548 3.731 3.548 3.731 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.603 3.691 ls_scf_post 1 4.0 0.000 0.000 3.679 3.686 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.202 3.648 make_images_sizes 222 9.7 0.000 0.000 1.011 3.610 mp_alltoall_i44 222 10.7 1.010 3.609 1.010 3.609 ls_scf_store_result 1 5.0 0.000 0.000 3.439 3.484 dbcsr_finalize 304 7.8 0.062 0.078 2.197 2.271 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.185 2.187 rebuild_ks_matrix 3 7.3 0.000 0.000 2.152 2.154 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.152 2.154 dbcsr_merge_all 275 8.9 0.473 0.522 2.045 2.109 dbcsr_data_release 10123 10.8 1.330 2.022 1.330 2.022 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.160000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3633.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5e06bd5b400cb74c2f9aca977c7c2ab82af71875_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 8.749638E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.078 0.098 108.590 108.591 qs_energies 1 2.0 0.000 0.000 107.176 107.197 ls_scf 1 3.0 0.000 0.000 104.204 104.224 dbcsr_multiply_generic 111 6.7 0.023 0.026 77.721 77.825 ls_scf_main 1 4.0 0.000 0.000 66.420 66.421 density_matrix_trs4 2 5.0 0.002 0.003 57.106 57.155 multiply_cannon 111 7.7 0.117 0.149 50.284 52.328 multiply_cannon_loop 111 8.7 0.067 0.070 46.811 47.580 ls_scf_init_scf 1 4.0 0.000 0.000 34.123 34.125 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.432 32.442 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.613 29.623 mp_waitall_1 4527 11.1 22.535 26.173 22.535 26.173 make_m2s 222 7.7 0.005 0.005 23.882 24.887 make_images 222 8.7 4.576 4.959 23.776 24.779 multiply_cannon_multrec 444 9.7 17.891 18.653 22.557 23.110 hybrid_alltoall_any 227 10.6 1.657 3.619 12.936 15.650 make_images_data 222 9.7 0.003 0.003 13.141 15.581 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.816 11.361 multiply_cannon_sync_h2d 444 9.7 8.843 8.892 8.843 8.892 arnoldi_extremal 4 6.8 0.000 0.000 7.331 7.342 arnoldi_normal_ev 4 7.8 0.003 0.009 7.331 7.342 build_subspace 16 8.4 0.026 0.036 6.792 6.802 dbcsr_matrix_vector_mult 304 9.0 0.016 0.032 5.423 5.578 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.260 5.351 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.030 5.283 dbcsr_matrix_vector_mult_local 304 10.0 4.954 5.246 4.956 5.248 dbcsr_mm_accdrv_process 1814 10.4 0.231 0.327 4.499 4.647 dbcsr_mm_accdrv_process_sort 1814 11.4 4.199 4.326 4.199 4.326 ls_scf_post 1 4.0 0.000 0.000 3.661 3.681 mp_allgather_i34 111 8.7 1.120 3.527 1.120 3.527 make_images_sizes 222 9.7 0.000 0.000 1.427 3.455 mp_alltoall_i44 222 10.7 1.427 3.455 1.427 3.455 ls_scf_store_result 1 5.0 0.000 0.000 3.414 3.420 calculate_norms 792 9.8 3.244 3.275 3.244 3.275 dbcsr_finalize 304 7.8 0.082 0.089 3.082 3.157 qs_energies_init_hamiltonians 1 3.0 0.032 0.037 2.941 2.942 dbcsr_merge_all 275 8.9 0.890 0.910 2.868 2.937 dbcsr_complete_redistribute 5 7.6 1.448 1.488 2.780 2.875 dbcsr_data_release 12724 10.6 2.337 2.869 2.337 2.869 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.621 2.623 rebuild_ks_matrix 3 7.3 0.000 0.000 2.555 2.557 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.555 2.557 matrix_ls_to_qs 2 6.0 0.000 0.000 2.415 2.516 dbcsr_sort_data 325 11.1 2.439 2.501 2.439 2.501 dbcsr_new_transposed 4 7.5 0.243 0.253 2.238 2.244 dbcsr_frobenius_norm 74 6.6 2.056 2.129 2.182 2.224 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.200 dbcsr_add_anytype 103 7.2 0.860 0.891 2.131 2.200 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.591000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6809.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 5e06bd5b400cb74c2f9aca977c7c2ab82af71875 Summary: empty Status: OK