=== This is the CP2K Performance-Test === Updating ac343fdab..1eb99f034 Fast-forward tools/toolchain/scripts/stage8/install_sirius.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 1eb99f0345e8f3b2962f17f4aadd5584dc421ad2 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 04.10.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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)) 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__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 += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/01 job id: 49875900 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/02 job id: 49875901 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/03 job id: 49875902 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/04 job id: 49875904 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/05 job id: 49875905 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/06 job id: 49875906 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/07 job id: 49875907 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/08 job id: 49875908 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/09 job id: 49875909 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/10 job id: 49875910 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/11 job id: 49875911 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/12 job id: 49875912 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/13 job id: 49875913 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/14 job id: 49875915 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/15 job id: 49875917 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/16 job id: 49875919 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/17 job id: 49875921 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/18 job id: 49875922 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/19 job id: 49875923 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/20 job id: 49875924 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/21 job id: 49875925 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/22 job id: 49875927 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/23 job id: 49875928 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/24 job id: 49875929 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/25 job id: 49875930 --- 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/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/26 job id: 49875931 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/27 job id: 49875932 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.034 134.143 134.143 farming_run 1 2.0 133.554 133.556 134.113 134.117 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.487827E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.024 114.652 114.652 qs_energies 1 2.0 0.000 0.000 114.462 114.464 mp2_main 1 3.0 0.000 0.000 112.366 112.369 mp2_gpw_main 1 4.0 0.034 0.046 111.376 111.378 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 92.421 92.768 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.697 55.044 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.166 41.078 46.129 get_2c_integrals 1 6.0 0.008 0.009 36.763 37.545 integrate_v_rspace 273 8.0 0.435 0.452 24.618 29.438 pw_transfer 6555 10.6 0.387 0.399 26.816 27.229 fft_wrap_pw1pw2 5465 11.4 0.046 0.049 25.445 25.931 grid_integrate_task_list 273 9.0 20.518 25.812 20.518 25.812 fft_wrap_pw1pw2_100 2178 12.4 1.097 1.281 23.006 23.477 compute_2c_integrals 1 7.0 0.002 0.002 19.020 19.026 rpa_ri_compute_en 1 5.0 0.019 0.022 18.843 19.018 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 18.578 18.741 mp2_eri_2c_integrate_gpw 1 9.0 2.381 2.416 18.577 18.741 cp_fm_cholesky_decompose 12 8.2 17.668 18.458 17.668 18.458 cholesky_decomp 1 7.0 0.000 0.000 16.577 17.363 fft3d_s 5443 13.4 16.121 16.523 16.143 16.546 ao_to_mo_and_store_B_mult_1 272 7.0 10.778 15.320 10.778 15.320 calculate_wavefunction 272 8.0 5.425 5.570 12.357 13.050 rpa_num_int 1 6.0 0.000 0.001 10.628 10.638 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.608 10.613 calc_mat_Q 8 8.0 0.000 0.000 9.468 9.557 contract_S_to_Q 8 9.0 0.000 0.000 8.891 8.977 calc_potential_gpw 544 9.5 0.005 0.005 8.255 8.638 parallel_gemm_fm 14 9.1 0.000 0.000 8.486 8.591 parallel_gemm_fm_cosma 14 10.1 8.486 8.591 8.486 8.591 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.064 8.273 potential_pw2rs 545 10.0 0.107 0.109 7.475 8.098 create_integ_mat 1 6.0 0.014 0.030 7.774 7.784 collocate_single_gaussian 272 10.0 0.039 0.042 7.318 7.567 array2fm 1 7.0 0.000 0.000 6.639 7.213 pw_scatter_s 2720 13.7 4.379 4.625 4.379 4.625 pw_gather_s 2722 13.2 3.447 3.922 3.447 3.922 array2fm_buffer_send 1 8.0 2.876 3.063 2.876 3.063 pw_poisson_solve 545 10.5 1.128 1.184 2.196 2.326 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.378095, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2805.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.038 421.939 421.940 farming_run 1 2.0 420.812 420.817 421.907 421.908 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.243230E+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 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 16427074. 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 0.011 0.031 217.098 217.098 qs_energies 1 2.0 0.000 0.000 216.851 216.862 scf_env_do_scf 1 3.0 0.000 0.000 114.982 114.982 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.079 114.086 rebuild_ks_matrix 4 6.0 0.000 0.000 114.078 114.084 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 114.078 114.084 hfx_ks_matrix 4 8.0 0.001 0.001 113.734 113.739 integrate_four_center 4 9.0 0.152 0.462 113.733 113.738 integrate_four_center_main 4 10.0 0.127 0.573 101.791 105.277 integrate_four_center_bin 265 11.0 101.664 105.053 101.664 105.053 mp2_main 1 3.0 0.000 0.000 101.529 101.539 mp2_gpw_main 1 4.0 0.048 0.077 100.691 100.703 init_scf_loop 1 4.0 0.000 0.000 96.380 96.380 mp2_ri_gpw_compute_in 1 5.0 0.063 0.064 73.888 75.006 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.803 54.918 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.160 41.495 46.484 integrate_v_rspace 95 8.0 0.394 0.566 27.963 32.808 pw_transfer 2240 10.6 0.147 0.167 29.496 29.956 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.502 28.978 mp2_ri_gpw_compute_en 1 5.0 0.055 0.065 26.656 28.634 grid_integrate_task_list 95 9.0 23.384 28.338 23.384 28.338 ao_to_mo_and_store_B_mult_1 91 7.0 10.634 26.959 10.634 26.959 fft_wrap_pw1pw2_100 730 12.4 1.279 1.448 26.252 26.738 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.828 1.882 24.892 24.901 get_2c_integrals 1 6.0 0.000 0.000 19.997 20.021 compute_2c_integrals 1 7.0 0.002 0.003 18.978 18.982 fft3d_s 1823 13.4 18.486 18.934 18.500 18.947 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.629 18.859 mp2_eri_2c_integrate_gpw 1 9.0 1.733 1.837 18.628 18.859 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.600 18.600 calculate_wavefunction 91 8.0 2.016 2.048 9.622 9.866 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.587 8.779 9.304 potential_pw2rs 186 10.0 0.033 0.035 8.404 8.895 local_gemm 172 8.0 8.222 8.749 8.222 8.749 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.151 8.449 mp2_ri_gpw_compute_en_comm 22 7.0 0.502 0.521 7.902 8.414 calc_potential_gpw 182 9.5 0.002 0.002 7.917 8.198 collocate_single_gaussian 91 10.0 0.017 0.024 7.754 7.931 mp_sync 37 10.5 3.930 7.057 3.930 7.057 integrate_four_center_load 4 10.0 0.000 0.000 6.751 6.755 hfx_load_balance 1 11.0 0.000 0.000 6.751 6.755 mp_sendrecv_dm3 2068 8.0 5.929 6.427 5.929 6.427 mp2_ri_gpw_compute_en_ener 172 7.0 6.338 6.410 6.338 6.410 pw_gather_s 912 13.2 4.475 5.010 4.475 5.010 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.691444, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1502.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_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 451.985408E+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 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. 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 0.006 0.027 52.060 52.061 qs_mol_dyn_low 1 2.0 0.003 0.003 51.855 51.863 qs_forces 11 3.9 0.002 0.002 51.787 51.788 qs_energies 11 4.9 0.001 0.001 50.322 50.338 scf_env_do_scf 11 5.9 0.001 0.001 43.623 43.623 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 41.638 41.639 dbcsr_multiply_generic 2286 12.5 0.096 0.104 32.532 33.137 qs_scf_new_mos 108 7.5 0.000 0.001 31.866 32.144 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.865 32.143 ot_scf_mini 108 9.5 0.002 0.002 30.259 30.441 multiply_cannon 2286 13.5 0.193 0.203 26.466 28.339 multiply_cannon_loop 2286 14.5 1.761 1.883 25.821 27.682 velocity_verlet 10 3.0 0.001 0.001 25.368 25.368 ot_mini 108 10.5 0.001 0.001 17.988 18.243 qs_ot_get_derivative 108 11.5 0.001 0.001 15.017 15.198 mp_waitall_1 245248 16.5 7.857 14.292 7.857 14.292 multiply_cannon_metrocomm3 54864 15.5 0.074 0.081 5.857 13.167 multiply_cannon_multrec 54864 15.5 3.783 5.930 8.186 11.766 qs_ot_get_p 119 10.4 0.001 0.001 7.732 7.980 rebuild_ks_matrix 119 8.3 0.000 0.000 7.709 7.843 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.709 7.842 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.778 6.896 dbcsr_mm_accdrv_process 76910 16.1 1.737 2.816 4.316 6.208 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.583 6.037 mp_sum_l 7287 12.8 4.160 5.954 4.160 5.954 multiply_cannon_sync_h2d 54864 15.5 5.301 5.827 5.301 5.827 init_scf_run 11 5.9 0.000 0.001 5.363 5.364 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.363 5.363 qs_ot_p2m_diag 50 11.0 0.004 0.006 4.966 5.009 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.419 4.520 sum_up_and_integrate 119 10.3 0.001 0.002 4.364 4.369 integrate_v_rspace 119 11.3 0.002 0.002 4.353 4.360 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.280 4.280 qs_rho_update_rho_low 119 7.7 0.000 0.001 3.996 4.108 calculate_rho_elec 119 8.7 0.011 0.016 3.995 4.107 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.095 4.095 cp_fm_redistribute_end 50 14.0 2.091 4.073 2.096 4.075 cp_fm_diag_elpa_base 50 14.0 1.973 3.968 1.977 3.975 jit_kernel_multiply 13 15.8 2.519 3.322 2.519 3.322 calculate_first_density_matrix 1 7.0 0.000 0.001 3.183 3.189 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.921 3.171 calculate_dm_sparse 119 9.5 0.000 0.001 3.042 3.171 apply_single 119 13.6 0.000 0.000 2.921 3.171 acc_transpose_blocks 54864 15.5 0.217 0.252 2.369 3.093 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.770 2.772 ot_diis_step 108 11.5 0.006 0.006 2.738 2.739 multiply_cannon_metrocomm1 54864 15.5 0.057 0.062 1.401 2.517 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.458 2.511 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.334 2.385 density_rs2pw 119 9.7 0.004 0.004 2.067 2.183 wfi_extrapolate 11 7.9 0.001 0.001 2.110 2.110 grid_integrate_task_list 119 12.3 2.015 2.102 2.015 2.102 init_scf_loop 11 6.9 0.000 0.000 1.968 1.968 mp_sum_d 4135 12.0 1.303 1.953 1.303 1.953 potential_pw2rs 119 12.3 0.004 0.004 1.750 1.761 pw_transfer 1439 11.6 0.052 0.060 1.589 1.657 make_m2s 4572 13.5 0.054 0.055 1.588 1.639 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.597 1.627 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.513 1.584 make_images 4572 14.5 0.134 0.139 1.505 1.556 mp_alltoall_d11v 2130 13.8 1.222 1.431 1.222 1.431 grid_collocate_task_list 119 9.7 1.357 1.425 1.357 1.425 transfer_rs2pw 487 10.6 0.005 0.006 1.318 1.409 acc_transpose_blocks_kernels 54864 16.5 0.238 0.373 0.978 1.408 acc_transpose_blocks_sync 164592 16.5 1.129 1.393 1.129 1.393 mp_waitany 12084 13.8 1.245 1.345 1.245 1.345 transfer_pw2rs 487 13.2 0.006 0.007 1.284 1.292 fft3d_ps 1201 14.6 0.369 0.474 1.226 1.289 fft_wrap_pw1pw2_140 487 13.2 0.142 0.156 1.176 1.248 dbcsr_dot_sd 1205 11.9 0.048 0.060 0.709 1.124 jit_kernel_transpose 5 15.5 0.740 1.060 0.740 1.060 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.061000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.272727, yerr=0.962091 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_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 489.422848E+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 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 509581. 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 0.007 0.025 40.144 40.145 qs_mol_dyn_low 1 2.0 0.003 0.005 39.940 39.948 qs_forces 11 3.9 0.002 0.003 39.879 39.879 qs_energies 11 4.9 0.001 0.001 38.208 38.213 scf_env_do_scf 11 5.9 0.001 0.001 32.306 32.306 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.919 29.919 dbcsr_multiply_generic 2286 12.5 0.101 0.104 23.211 23.552 qs_scf_new_mos 108 7.5 0.001 0.001 21.385 21.609 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.385 21.609 ot_scf_mini 108 9.5 0.002 0.003 20.477 20.636 multiply_cannon 2286 13.5 0.210 0.218 18.090 19.514 multiply_cannon_loop 2286 14.5 1.188 1.251 16.934 18.460 velocity_verlet 10 3.0 0.001 0.002 18.186 18.187 ot_mini 108 10.5 0.001 0.001 12.988 13.215 mp_waitall_1 200699 16.5 6.180 11.554 6.180 11.554 qs_ot_get_derivative 108 11.5 0.001 0.001 10.629 10.788 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.352 10.333 multiply_cannon_multrec 27432 15.5 1.938 4.634 6.661 9.639 rebuild_ks_matrix 119 8.3 0.000 0.000 6.725 6.862 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.725 6.862 dbcsr_mm_accdrv_process 47894 16.0 3.811 6.229 4.643 6.744 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.908 6.033 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.912 4.838 qs_ot_get_p 119 10.4 0.001 0.001 4.573 4.788 init_scf_run 11 5.9 0.000 0.001 4.709 4.709 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.709 4.709 mp_sum_l 7287 12.8 2.094 4.102 2.094 4.102 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.079 4.060 apply_single 119 13.6 0.000 0.000 3.079 4.060 sum_up_and_integrate 119 10.3 0.001 0.001 3.665 3.672 integrate_v_rspace 119 11.3 0.002 0.003 3.651 3.658 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.569 3.599 calculate_rho_elec 119 8.7 0.021 0.024 3.569 3.598 calculate_first_density_matrix 1 7.0 0.000 0.000 3.202 3.203 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.943 2.961 make_m2s 4572 13.5 0.052 0.054 2.578 2.911 make_images 4572 14.5 0.207 0.245 2.489 2.824 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.548 2.549 calculate_dm_sparse 119 9.5 0.000 0.001 2.384 2.474 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.307 2.390 init_scf_loop 11 6.9 0.000 0.000 2.368 2.368 ot_diis_step 108 11.5 0.011 0.011 2.312 2.312 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.238 2.238 acc_transpose_blocks 27432 15.5 0.113 0.118 1.653 2.227 cp_fm_redistribute_end 50 14.0 1.135 2.215 1.138 2.217 multiply_cannon_sync_h2d 27432 15.5 1.693 2.201 1.693 2.201 cp_fm_diag_elpa_base 50 14.0 1.048 2.131 1.076 2.168 jit_kernel_multiply 9 16.5 0.773 2.158 0.773 2.158 grid_integrate_task_list 119 12.3 1.844 1.960 1.844 1.960 density_rs2pw 119 9.7 0.004 0.004 1.870 1.942 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.938 1.940 pw_transfer 1439 11.6 0.065 0.069 1.817 1.850 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.746 1.788 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.725 1.761 make_images_data 4572 15.5 0.047 0.054 1.227 1.593 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.582 1.592 prepare_preconditioner 11 7.9 0.000 0.000 1.489 1.516 make_preconditioner 11 8.9 0.000 0.000 1.489 1.516 wfi_extrapolate 11 7.9 0.001 0.001 1.449 1.449 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.388 1.445 hybrid_alltoall_any 4725 16.4 0.054 0.116 1.091 1.436 potential_pw2rs 119 12.3 0.006 0.006 1.408 1.412 fft_wrap_pw1pw2_140 487 13.2 0.159 0.168 1.376 1.411 fft3d_ps 1201 14.6 0.518 0.572 1.362 1.389 grid_collocate_task_list 119 9.7 1.288 1.346 1.288 1.346 mp_alltoall_d11v 2130 13.8 1.188 1.309 1.188 1.309 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.198 1.247 acc_transpose_blocks_kernels 27432 16.5 0.189 0.277 0.696 1.149 transfer_rs2pw 487 10.6 0.005 0.005 1.057 1.139 multiply_cannon_metrocomm1 27432 15.5 0.032 0.035 0.652 1.124 mp_sum_d 4135 12.0 0.534 1.016 0.534 1.016 parallel_gemm_fm 81 9.0 0.000 0.000 0.620 0.980 parallel_gemm_fm_cosma 81 10.0 0.620 0.979 0.620 0.979 mp_allgather_i34 2286 14.5 0.577 0.967 0.577 0.967 acc_transpose_blocks_sync 82296 16.5 0.817 0.940 0.817 0.940 cp_fm_cholesky_decompose 22 10.9 0.771 0.937 0.771 0.937 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.935 0.936 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.872 0.885 transfer_pw2rs 487 13.2 0.004 0.006 0.879 0.885 make_basis_sm 11 9.8 0.000 0.000 0.877 0.878 jit_kernel_transpose 5 15.5 0.507 0.875 0.507 0.875 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.145000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.727273, yerr=1.542778 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_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 522.498048E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.028 33.005 33.006 qs_mol_dyn_low 1 2.0 0.003 0.003 32.763 32.770 qs_forces 11 3.9 0.002 0.002 32.693 32.693 qs_energies 11 4.9 0.001 0.001 31.157 31.158 scf_env_do_scf 11 5.9 0.000 0.001 25.508 25.508 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.109 23.110 dbcsr_multiply_generic 2286 12.5 0.097 0.098 18.194 18.280 multiply_cannon 2286 13.5 0.195 0.199 14.986 15.709 qs_scf_new_mos 108 7.5 0.001 0.001 15.539 15.553 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.538 15.552 velocity_verlet 10 3.0 0.001 0.002 15.000 15.002 multiply_cannon_loop 2286 14.5 0.853 0.891 14.224 14.967 ot_scf_mini 108 9.5 0.002 0.002 14.783 14.795 ot_mini 108 10.5 0.001 0.001 9.274 9.285 multiply_cannon_multrec 18288 15.5 1.868 2.841 7.713 7.965 qs_ot_get_derivative 108 11.5 0.001 0.001 7.799 7.809 dbcsr_mm_accdrv_process 38222 16.0 5.659 6.616 5.751 6.680 rebuild_ks_matrix 119 8.3 0.000 0.000 5.904 5.916 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.904 5.916 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.197 5.208 init_scf_run 11 5.9 0.000 0.001 4.499 4.499 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.499 4.499 mp_waitall_1 158411 16.6 2.743 3.987 2.743 3.987 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.083 3.708 sum_up_and_integrate 119 10.3 0.001 0.001 3.493 3.497 integrate_v_rspace 119 11.3 0.002 0.003 3.480 3.486 qs_ot_get_p 119 10.4 0.001 0.001 3.279 3.301 calculate_first_density_matrix 1 7.0 0.000 0.000 3.285 3.286 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.212 3.217 calculate_rho_elec 119 8.7 0.031 0.031 3.212 3.217 calculate_dm_sparse 119 9.5 0.000 0.001 2.746 2.757 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.624 2.730 init_scf_loop 11 6.9 0.000 0.000 2.382 2.384 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.021 2.333 apply_single 119 13.6 0.000 0.000 2.021 2.333 acc_transpose_blocks 18288 15.5 0.079 0.082 2.050 2.329 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.191 2.198 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.086 2.089 make_m2s 4572 13.5 0.044 0.045 1.764 1.912 grid_integrate_task_list 119 12.3 1.801 1.906 1.801 1.906 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.896 1.896 density_rs2pw 119 9.7 0.004 0.004 1.779 1.887 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.864 1.865 make_images 4572 14.5 0.193 0.205 1.678 1.825 pw_transfer 1439 11.6 0.065 0.069 1.797 1.804 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.716 1.721 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.704 1.714 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.698 1.702 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.654 1.666 cp_fm_diag_elpa_base 50 14.0 1.632 1.644 1.653 1.665 prepare_preconditioner 11 7.9 0.000 0.000 1.638 1.641 make_preconditioner 11 8.9 0.000 0.000 1.638 1.641 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.500 1.581 ot_diis_step 108 11.5 0.011 0.012 1.461 1.461 acc_transpose_blocks_kernels 18288 16.5 0.219 0.226 1.211 1.442 mp_sum_l 7287 12.8 1.051 1.401 1.051 1.401 fft_wrap_pw1pw2_140 487 13.2 0.211 0.216 1.384 1.394 grid_collocate_task_list 119 9.7 1.233 1.314 1.233 1.314 potential_pw2rs 119 12.3 0.007 0.009 1.294 1.296 fft3d_ps 1201 14.6 0.527 0.547 1.273 1.280 multiply_cannon_sync_h2d 18288 15.5 1.042 1.230 1.042 1.230 jit_kernel_transpose 5 15.6 0.992 1.217 0.992 1.217 wfi_extrapolate 11 7.9 0.001 0.001 1.162 1.162 transfer_rs2pw 487 10.6 0.005 0.005 0.945 1.058 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.937 0.954 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.931 0.931 make_images_data 4572 15.5 0.047 0.051 0.756 0.926 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.650 0.869 acc_transpose_blocks_sync 54864 16.5 0.740 0.816 0.740 0.816 transfer_pw2rs 487 13.2 0.004 0.004 0.781 0.782 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.777 0.778 mp_alltoall_d11v 2130 13.8 0.654 0.757 0.654 0.757 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.673 0.729 cp_fm_cholesky_invert 11 10.9 0.684 0.688 0.684 0.688 mp_alltoall_z22v 1201 16.6 0.599 0.663 0.599 0.663 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.006000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.454545, yerr=2.061052 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 556.408832E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.026 36.548 36.549 qs_mol_dyn_low 1 2.0 0.003 0.003 36.344 36.353 qs_forces 11 3.9 0.002 0.003 36.229 36.230 qs_energies 11 4.9 0.001 0.001 34.524 34.529 scf_env_do_scf 11 5.9 0.000 0.001 28.935 28.937 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.627 25.627 dbcsr_multiply_generic 2286 12.5 0.101 0.103 20.352 20.491 velocity_verlet 10 3.0 0.001 0.002 18.301 18.305 qs_scf_new_mos 108 7.5 0.001 0.001 17.625 17.693 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.624 17.692 multiply_cannon 2286 13.5 0.218 0.227 16.667 17.278 ot_scf_mini 108 9.5 0.002 0.003 16.595 16.652 multiply_cannon_loop 2286 14.5 1.507 1.622 15.721 16.159 ot_mini 108 10.5 0.001 0.001 10.228 10.291 multiply_cannon_multrec 27432 15.5 2.447 3.143 9.330 9.660 qs_ot_get_derivative 108 11.5 0.001 0.001 8.371 8.427 dbcsr_mm_accdrv_process 47916 15.9 6.208 7.691 6.779 8.092 rebuild_ks_matrix 119 8.3 0.000 0.000 6.287 6.340 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.287 6.339 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.568 5.613 init_scf_run 11 5.9 0.000 0.001 4.266 4.267 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.266 4.267 qs_ot_get_p 119 10.4 0.001 0.001 3.522 3.601 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.071 3.511 sum_up_and_integrate 119 10.3 0.001 0.001 3.463 3.472 integrate_v_rspace 119 11.3 0.002 0.003 3.451 3.461 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.307 3.335 calculate_rho_elec 119 8.7 0.040 0.046 3.307 3.335 init_scf_loop 11 6.9 0.000 0.000 3.290 3.290 calculate_first_density_matrix 1 7.0 0.000 0.000 2.887 2.889 acc_transpose_blocks 27432 15.5 0.119 0.122 2.746 2.875 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.147 2.576 apply_single 119 13.6 0.000 0.000 2.147 2.576 calculate_dm_sparse 119 9.5 0.000 0.000 2.449 2.503 prepare_preconditioner 11 7.9 0.000 0.000 2.470 2.477 make_preconditioner 11 8.9 0.000 0.000 2.470 2.477 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.083 2.402 make_m2s 4572 13.5 0.054 0.056 2.224 2.358 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.290 2.318 make_images 4572 14.5 0.275 0.336 2.116 2.248 mp_waitall_1 137007 16.6 1.641 2.205 1.641 2.205 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.170 2.170 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.145 2.158 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.001 2.011 pw_transfer 1439 11.6 0.065 0.069 1.916 1.949 grid_integrate_task_list 119 12.3 1.821 1.925 1.821 1.925 density_rs2pw 119 9.7 0.004 0.004 1.763 1.896 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.824 1.860 ot_diis_step 108 11.5 0.012 0.013 1.812 1.812 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.751 1.751 mp_sum_l 7287 12.8 1.031 1.642 1.031 1.642 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.546 1.558 fft_wrap_pw1pw2_140 487 13.2 0.246 0.258 1.507 1.544 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.518 1.529 cp_fm_diag_elpa_base 50 14.0 1.484 1.503 1.516 1.527 acc_transpose_blocks_sync 82296 16.5 1.399 1.521 1.399 1.521 jit_kernel_multiply 6 16.2 0.504 1.414 0.504 1.414 fft3d_ps 1201 14.6 0.555 0.604 1.343 1.366 grid_collocate_task_list 119 9.7 1.249 1.340 1.249 1.340 wfi_extrapolate 11 7.9 0.001 0.001 1.326 1.327 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.281 1.300 potential_pw2rs 119 12.3 0.009 0.010 1.268 1.271 acc_transpose_blocks_kernels 27432 16.5 0.272 0.279 1.199 1.221 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.689 1.193 cp_fm_upper_to_full 72 14.2 0.808 1.146 0.808 1.146 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.067 1.068 transfer_rs2pw 487 10.6 0.004 0.005 0.876 1.012 dbcsr_complete_redistribute 329 12.2 0.121 0.148 0.736 1.010 make_images_data 4572 15.5 0.047 0.051 0.819 0.948 jit_kernel_transpose 5 15.6 0.927 0.943 0.927 0.943 hybrid_alltoall_any 4725 16.4 0.066 0.157 0.697 0.898 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.801 0.879 mp_alltoall_d11v 2130 13.8 0.727 0.836 0.727 0.836 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.552 0.823 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.808 0.813 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.549000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.363636, yerr=3.574473 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_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 627.597312E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.028 28.323 28.324 qs_mol_dyn_low 1 2.0 0.003 0.003 28.153 28.161 qs_forces 11 3.9 0.002 0.003 28.079 28.080 qs_energies 11 4.9 0.001 0.001 26.373 26.376 scf_env_do_scf 11 5.9 0.000 0.001 21.695 21.695 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.285 19.286 velocity_verlet 10 3.0 0.002 0.002 14.581 14.585 dbcsr_multiply_generic 2286 12.5 0.095 0.099 13.077 13.144 qs_scf_new_mos 108 7.5 0.001 0.001 11.728 11.751 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.727 11.750 multiply_cannon 2286 13.5 0.223 0.235 10.500 11.049 ot_scf_mini 108 9.5 0.002 0.002 11.016 11.039 multiply_cannon_loop 2286 14.5 0.642 0.666 9.585 9.773 ot_mini 108 10.5 0.001 0.001 6.332 6.361 multiply_cannon_multrec 9144 15.5 1.687 1.965 6.018 6.257 rebuild_ks_matrix 119 8.3 0.000 0.000 5.711 5.738 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.710 5.737 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.074 5.098 qs_ot_get_derivative 108 11.5 0.001 0.001 5.050 5.074 dbcsr_mm_accdrv_process 12550 15.8 3.717 4.237 4.222 4.321 sum_up_and_integrate 119 10.3 0.001 0.001 3.400 3.403 integrate_v_rspace 119 11.3 0.003 0.005 3.389 3.393 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.302 3.309 calculate_rho_elec 119 8.7 0.060 0.061 3.301 3.308 init_scf_run 11 5.9 0.000 0.001 3.248 3.248 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.247 3.247 qs_ot_get_p 119 10.4 0.001 0.001 2.808 2.843 init_scf_loop 11 6.9 0.000 0.000 2.389 2.391 calculate_first_density_matrix 1 7.0 0.000 0.000 2.122 2.123 pw_transfer 1439 11.6 0.066 0.068 1.945 1.953 make_m2s 4572 13.5 0.034 0.036 1.776 1.941 grid_integrate_task_list 119 12.3 1.866 1.936 1.866 1.936 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.851 1.861 make_images 4572 14.5 0.271 0.304 1.687 1.850 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.835 1.838 mp_waitall_1 115863 16.7 1.341 1.835 1.341 1.835 calculate_dm_sparse 119 9.5 0.000 0.000 1.792 1.813 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.750 1.772 density_rs2pw 119 9.7 0.003 0.004 1.676 1.772 prepare_preconditioner 11 7.9 0.000 0.000 1.674 1.679 make_preconditioner 11 8.9 0.000 0.000 1.674 1.679 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.564 1.592 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.591 1.591 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.549 1.550 fft_wrap_pw1pw2_140 487 13.2 0.322 0.330 1.528 1.539 acc_transpose_blocks 9144 15.5 0.041 0.043 1.453 1.495 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.391 1.400 grid_collocate_task_list 119 9.7 1.296 1.392 1.296 1.392 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.327 1.339 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.299 1.307 cp_fm_diag_elpa_base 50 14.0 1.273 1.290 1.298 1.306 fft3d_ps 1201 14.6 0.558 0.569 1.266 1.275 ot_diis_step 108 11.5 0.013 0.013 1.270 1.271 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.209 1.265 apply_single 119 13.6 0.000 0.000 1.209 1.265 jit_kernel_multiply 7 15.6 0.465 1.248 0.465 1.248 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.221 1.222 potential_pw2rs 119 12.3 0.010 0.011 1.213 1.215 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.176 1.182 wfi_extrapolate 11 7.9 0.001 0.001 1.076 1.076 hybrid_alltoall_any 4725 16.4 0.065 0.174 0.769 1.005 make_images_data 4572 15.5 0.042 0.046 0.769 0.951 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.878 0.932 mp_alltoall_d11v 2130 13.8 0.791 0.895 0.791 0.895 transfer_rs2pw 487 10.6 0.004 0.005 0.769 0.851 cp_fm_cholesky_invert 11 10.9 0.834 0.836 0.834 0.836 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.816 0.822 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.758 0.760 acc_transpose_blocks_sync 27432 16.5 0.721 0.758 0.721 0.758 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 0.369 0.749 qs_env_update_s_mstruct 11 6.9 0.001 0.002 0.668 0.717 acc_transpose_blocks_kernels 9144 16.5 0.119 0.122 0.675 0.679 mp_allgather_i34 2286 14.5 0.233 0.654 0.233 0.654 transfer_pw2rs 487 13.2 0.003 0.004 0.649 0.651 mp_alltoall_z22v 1201 16.6 0.582 0.609 0.582 0.609 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.324000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=594.818182, yerr=6.117310 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_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 770.760704E+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 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.057 0.061 43.800 43.801 qs_mol_dyn_low 1 2.0 0.003 0.003 43.532 43.585 qs_forces 11 3.9 0.010 0.015 43.466 43.466 qs_energies 11 4.9 0.001 0.001 41.465 41.476 scf_env_do_scf 11 5.9 0.001 0.001 34.445 34.445 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.768 26.769 velocity_verlet 10 3.0 0.002 0.002 23.586 23.592 dbcsr_multiply_generic 2286 12.5 0.103 0.107 20.668 20.969 multiply_cannon 2286 13.5 0.298 0.310 16.855 17.919 qs_scf_new_mos 108 7.5 0.001 0.001 17.538 17.638 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.537 17.637 multiply_cannon_loop 2286 14.5 0.855 0.869 15.618 16.718 ot_scf_mini 108 9.5 0.002 0.002 16.387 16.482 ot_mini 108 10.5 0.001 0.001 10.173 10.286 multiply_cannon_multrec 9144 15.5 3.681 4.830 9.878 10.120 qs_ot_get_derivative 108 11.5 0.001 0.001 8.148 8.243 init_scf_loop 11 6.9 0.000 0.000 7.651 7.654 dbcsr_mm_accdrv_process 12550 15.8 5.260 6.839 6.057 7.356 rebuild_ks_matrix 119 8.3 0.000 0.001 6.989 7.126 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.989 7.125 prepare_preconditioner 11 7.9 0.000 0.000 6.707 6.721 make_preconditioner 11 8.9 0.000 0.000 6.707 6.721 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.266 6.601 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.297 6.421 init_scf_run 11 5.9 0.000 0.001 4.978 4.978 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.978 4.978 cp_fm_upper_to_full 72 14.2 3.156 4.542 3.156 4.542 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.002 4.397 calculate_rho_elec 119 8.7 0.118 0.121 4.002 4.397 mp_waitall_1 94719 16.7 2.930 3.921 2.930 3.921 sum_up_and_integrate 119 10.3 0.001 0.001 3.714 3.720 integrate_v_rspace 119 11.3 0.004 0.004 3.704 3.709 calculate_first_density_matrix 1 7.0 0.000 0.000 3.499 3.557 qs_ot_get_p 119 10.4 0.001 0.001 3.306 3.441 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.918 3.342 calculate_dm_sparse 119 9.5 0.000 0.000 2.944 3.014 dbcsr_complete_redistribute 329 12.2 0.286 0.291 1.957 2.778 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.198 2.639 apply_single 119 13.6 0.000 0.000 2.198 2.639 pw_transfer 1439 11.6 0.069 0.069 2.546 2.550 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.513 2.514 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.663 2.470 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.447 2.452 make_m2s 4572 13.5 0.037 0.038 2.279 2.423 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.651 2.387 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.265 2.307 make_images 4572 14.5 0.357 0.391 2.159 2.302 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.436 2.239 mp_alltoall_i22 627 13.8 1.412 2.221 1.412 2.221 grid_integrate_task_list 119 12.3 2.073 2.099 2.073 2.099 fft_wrap_pw1pw2_140 487 13.2 0.575 0.579 2.072 2.079 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.024 2.074 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.954 2.018 ot_diis_step 108 11.5 0.014 0.014 2.001 2.001 density_rs2pw 119 9.7 0.003 0.004 1.990 2.000 acc_transpose_blocks 9144 15.5 0.044 0.044 1.853 1.958 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.944 1.946 mp_sum_l 7287 12.8 0.986 1.860 0.986 1.860 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.764 1.772 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.593 1.594 mp_alltoall_d11v 2130 13.8 1.180 1.574 1.180 1.574 fft3d_ps 1201 14.6 0.592 0.602 1.544 1.549 grid_collocate_task_list 119 9.7 1.518 1.537 1.518 1.537 cp_fm_cholesky_invert 11 10.9 1.383 1.386 1.383 1.386 wfi_extrapolate 11 7.9 0.001 0.001 1.361 1.361 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.299 1.317 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.314 1.314 cp_fm_diag_elpa_base 50 14.0 1.169 1.221 1.312 1.312 potential_pw2rs 119 12.3 0.014 0.014 1.235 1.236 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.025 1.224 jit_kernel_multiply 5 15.4 0.768 1.178 0.768 1.178 acc_transpose_blocks_sync 27432 16.5 1.091 1.118 1.091 1.118 make_images_data 4572 15.5 0.046 0.049 0.945 1.117 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.090 1.104 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.001 1.042 qs_create_task_list 11 7.9 0.000 0.000 0.929 0.939 generate_qs_task_list 11 8.9 0.367 0.388 0.928 0.939 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.901 0.914 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.801000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=722.454545, yerr=14.895781 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 500.764672E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1591969. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.028 81.273 81.275 qs_mol_dyn_low 1 2.0 0.003 0.003 80.920 80.942 qs_forces 11 3.9 0.003 0.003 80.811 80.812 qs_energies 11 4.9 0.001 0.001 77.951 77.968 scf_env_do_scf 11 5.9 0.001 0.001 68.992 68.996 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 63.607 63.609 dbcsr_multiply_generic 2055 12.4 0.106 0.107 51.015 51.230 qs_scf_new_mos 99 7.5 0.000 0.001 46.798 46.936 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.798 46.936 ot_scf_mini 99 9.5 0.002 0.002 44.437 44.547 multiply_cannon 2055 13.4 0.186 0.190 42.973 43.651 velocity_verlet 10 3.0 0.002 0.003 42.791 42.794 multiply_cannon_loop 2055 14.4 1.773 1.813 42.033 42.724 ot_mini 99 10.5 0.001 0.001 26.627 26.730 qs_ot_get_derivative 99 11.5 0.001 0.001 19.815 19.945 multiply_cannon_multrec 49320 15.4 11.414 12.102 17.587 18.294 rebuild_ks_matrix 110 8.3 0.000 0.000 14.365 14.465 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.364 14.464 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.595 12.687 mp_waitall_1 220248 16.4 10.712 11.520 10.712 11.520 multiply_cannon_sync_h2d 49320 15.4 9.643 10.142 9.643 10.142 qs_ot_get_p 110 10.4 0.001 0.001 9.431 9.555 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.944 8.436 multiply_cannon_metrocomm3 49320 15.4 0.083 0.085 6.602 7.755 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.152 7.593 apply_single 110 13.6 0.000 0.000 7.152 7.593 init_scf_run 11 5.9 0.000 0.001 6.932 6.932 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.931 6.932 sum_up_and_integrate 110 10.3 0.002 0.003 6.906 6.919 integrate_v_rspace 110 11.3 0.003 0.003 6.880 6.899 dbcsr_mm_accdrv_process 87628 16.1 3.039 3.130 6.042 6.716 ot_diis_step 99 11.5 0.006 0.006 6.564 6.564 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.305 6.322 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.174 6.301 calculate_rho_elec 110 8.6 0.020 0.024 6.174 6.300 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.532 5.532 init_scf_loop 11 6.9 0.000 0.000 5.359 5.359 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.202 5.268 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.978 5.005 cp_fm_diag_elpa_base 48 14.0 4.965 4.993 4.976 5.004 mp_sum_l 6594 12.7 3.726 4.518 3.726 4.518 calculate_dm_sparse 110 9.5 0.001 0.001 4.004 4.119 wfi_extrapolate 11 7.9 0.001 0.001 3.989 3.989 make_m2s 4110 13.4 0.061 0.065 3.857 3.958 make_images 4110 14.4 0.178 0.190 3.761 3.866 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.517 3.520 grid_integrate_task_list 110 12.3 3.271 3.429 3.271 3.429 density_rs2pw 110 9.6 0.004 0.004 3.246 3.428 multiply_cannon_metrocomm1 49320 15.4 0.064 0.067 2.326 3.333 prepare_preconditioner 11 7.9 0.000 0.000 3.219 3.248 make_preconditioner 11 8.9 0.000 0.000 3.219 3.248 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.188 3.235 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.092 3.147 jit_kernel_multiply 13 15.9 2.722 3.126 2.722 3.126 pw_transfer 1331 11.6 0.055 0.067 2.998 3.077 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.016 3.064 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 2.909 2.992 calculate_first_density_matrix 1 7.0 0.000 0.000 2.845 2.849 acc_transpose_blocks 49320 15.4 0.211 0.219 2.789 2.839 fft_wrap_pw1pw2_140 451 13.1 0.390 0.430 2.505 2.592 potential_pw2rs 110 12.3 0.005 0.006 2.504 2.518 mp_alltoall_d11v 2046 13.8 1.983 2.430 1.983 2.430 grid_collocate_task_list 110 9.6 2.163 2.271 2.163 2.271 fft3d_ps 1111 14.6 0.794 0.885 2.204 2.259 mp_waitany 14300 13.8 1.808 2.026 1.808 2.026 transfer_rs2pw 451 10.6 0.005 0.006 1.834 2.008 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.925 1.952 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.899 1.926 make_images_data 4110 15.4 0.044 0.047 1.767 1.893 mp_sum_d 3889 11.9 1.374 1.839 1.374 1.839 hybrid_alltoall_any 4261 16.3 0.085 0.483 1.534 1.792 cp_fm_cholesky_invert 11 10.9 1.745 1.749 1.745 1.749 transfer_pw2rs 451 13.1 0.006 0.007 1.693 1.702 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.275000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.545455, yerr=2.189381 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 590.880768E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 1389100. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.027 68.553 68.554 qs_mol_dyn_low 1 2.0 0.003 0.003 68.290 68.309 qs_forces 11 3.9 0.003 0.003 68.221 68.223 qs_energies 11 4.9 0.001 0.002 64.909 64.913 scf_env_do_scf 11 5.9 0.000 0.001 56.162 56.165 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.701 48.702 dbcsr_multiply_generic 2055 12.4 0.115 0.118 38.172 38.354 velocity_verlet 10 3.0 0.001 0.002 35.789 35.790 qs_scf_new_mos 99 7.5 0.001 0.001 32.932 33.078 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.931 33.077 multiply_cannon 2055 13.4 0.224 0.242 31.635 32.831 ot_scf_mini 99 9.5 0.003 0.003 31.266 31.398 multiply_cannon_loop 2055 14.4 1.163 1.189 30.376 31.234 ot_mini 99 10.5 0.001 0.001 18.298 18.435 multiply_cannon_multrec 24660 15.4 6.985 8.523 14.136 15.715 rebuild_ks_matrix 110 8.3 0.000 0.001 13.482 13.585 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.482 13.585 qs_ot_get_derivative 99 11.5 0.001 0.001 12.546 12.679 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.861 11.950 mp_waitall_1 176588 16.5 7.451 9.894 7.451 9.894 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.098 7.852 multiply_cannon_sync_h2d 24660 15.4 6.406 7.534 6.406 7.534 init_scf_loop 11 6.9 0.000 0.000 7.426 7.427 dbcsr_mm_accdrv_process 52282 16.1 5.647 6.396 6.979 7.308 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.438 7.073 apply_single 110 13.6 0.000 0.001 6.438 7.072 qs_ot_get_p 110 10.4 0.001 0.001 6.310 6.467 init_scf_run 11 5.9 0.000 0.001 6.387 6.388 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.387 6.387 sum_up_and_integrate 110 10.3 0.001 0.002 6.300 6.314 integrate_v_rspace 110 11.3 0.003 0.003 6.273 6.287 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.838 5.847 calculate_rho_elec 110 8.6 0.039 0.047 5.837 5.846 ot_diis_step 99 11.5 0.010 0.010 5.705 5.705 prepare_preconditioner 11 7.9 0.000 0.000 5.424 5.446 make_preconditioner 11 8.9 0.000 0.000 5.424 5.446 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.688 5.424 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.013 5.168 make_m2s 4110 13.4 0.057 0.060 4.187 4.661 make_images 4110 14.4 0.411 0.467 4.076 4.544 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.448 4.468 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.987 3.987 pw_transfer 1331 11.6 0.066 0.072 3.513 3.648 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.547 3.548 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.406 3.544 wfi_extrapolate 11 7.9 0.001 0.001 3.504 3.504 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.431 3.445 cp_fm_diag_elpa_base 48 14.0 3.386 3.404 3.428 3.441 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.271 3.354 grid_integrate_task_list 110 12.3 3.176 3.313 3.176 3.313 density_rs2pw 110 9.6 0.004 0.004 3.084 3.312 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.209 3.256 fft_wrap_pw1pw2_140 451 13.1 0.460 0.474 2.910 3.052 calculate_dm_sparse 110 9.5 0.001 0.001 2.996 3.025 calculate_first_density_matrix 1 7.0 0.000 0.000 2.787 2.789 hybrid_alltoall_any 4261 16.3 0.106 0.451 1.988 2.753 make_images_data 4110 15.4 0.049 0.053 2.269 2.738 fft3d_ps 1111 14.6 1.114 1.342 2.501 2.645 cp_fm_cholesky_invert 11 10.9 2.502 2.508 2.502 2.508 grid_collocate_task_list 110 9.6 2.170 2.321 2.170 2.321 mp_sum_l 6594 12.7 1.718 2.317 1.718 2.317 jit_kernel_multiply 11 16.4 0.965 2.237 0.965 2.237 acc_transpose_blocks 24660 15.4 0.116 0.119 2.187 2.233 potential_pw2rs 110 12.3 0.009 0.009 2.178 2.186 mp_alltoall_d11v 2046 13.8 1.720 2.087 1.720 2.087 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.953 1.971 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.854 1.855 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.776 1.787 multiply_cannon_metrocomm4 22605 15.4 0.080 0.084 0.778 1.599 mp_allgather_i34 2055 14.4 0.566 1.592 0.566 1.592 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.566 1.575 transfer_rs2pw 451 10.6 0.006 0.007 1.385 1.549 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.390 1.493 mp_irecv_dv 57340 16.2 0.646 1.478 0.646 1.478 dbcsr_complete_redistribute 325 12.2 0.241 0.316 1.110 1.378 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.554000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.727273, yerr=6.850228 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_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 664.379392E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.029 63.621 63.622 qs_mol_dyn_low 1 2.0 0.003 0.003 63.362 63.372 qs_forces 11 3.9 0.003 0.003 63.276 63.279 qs_energies 11 4.9 0.001 0.002 60.035 60.041 scf_env_do_scf 11 5.9 0.000 0.001 51.359 51.360 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.180 42.181 velocity_verlet 10 3.0 0.001 0.001 34.032 34.034 dbcsr_multiply_generic 2055 12.4 0.108 0.113 29.933 30.152 qs_scf_new_mos 99 7.5 0.001 0.001 26.933 27.014 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.932 27.013 ot_scf_mini 99 9.5 0.002 0.003 25.660 25.761 multiply_cannon 2055 13.4 0.214 0.225 23.146 24.338 multiply_cannon_loop 2055 14.4 0.813 0.835 21.890 22.976 ot_mini 99 10.5 0.001 0.001 14.187 14.292 rebuild_ks_matrix 110 8.3 0.000 0.000 12.495 12.621 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.494 12.620 multiply_cannon_multrec 16440 15.4 3.742 4.697 10.507 11.406 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.998 11.115 mp_waitall_1 139946 16.5 7.260 10.211 7.260 10.211 qs_ot_get_derivative 99 11.5 0.001 0.001 9.753 9.855 init_scf_loop 11 6.9 0.000 0.000 9.133 9.135 prepare_preconditioner 11 7.9 0.000 0.000 7.355 7.371 make_preconditioner 11 8.9 0.000 0.000 7.355 7.370 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.204 7.060 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.679 7.037 dbcsr_mm_accdrv_process 34862 16.1 5.880 6.321 6.607 6.740 qs_ot_get_p 110 10.4 0.001 0.001 6.562 6.698 sum_up_and_integrate 110 10.3 0.001 0.002 6.581 6.596 integrate_v_rspace 110 11.3 0.003 0.003 6.555 6.570 init_scf_run 11 5.9 0.000 0.001 6.228 6.229 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.228 6.228 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.105 6.119 calculate_rho_elec 110 8.6 0.059 0.059 6.105 6.118 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.903 5.364 apply_single 110 13.6 0.000 0.000 4.903 5.364 make_m2s 4110 13.4 0.050 0.052 4.660 5.036 make_images 4110 14.4 0.399 0.523 4.544 4.920 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.737 4.741 ot_diis_step 99 11.5 0.011 0.011 4.402 4.403 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.320 4.320 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.329 3.948 multiply_cannon_sync_h2d 16440 15.4 3.231 3.794 3.231 3.794 pw_transfer 1331 11.6 0.066 0.075 3.762 3.783 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.655 3.679 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.658 3.669 cp_fm_diag_elpa_base 48 14.0 3.583 3.621 3.656 3.667 density_rs2pw 110 9.6 0.004 0.005 3.343 3.499 grid_integrate_task_list 110 12.3 3.172 3.445 3.172 3.445 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.261 3.262 wfi_extrapolate 11 7.9 0.001 0.001 3.120 3.120 cp_fm_cholesky_invert 11 10.9 3.082 3.089 3.082 3.089 fft_wrap_pw1pw2_140 451 13.1 0.577 0.584 3.054 3.078 make_images_data 4110 15.4 0.045 0.050 2.566 3.075 hybrid_alltoall_any 4261 16.3 0.108 0.380 2.269 3.038 calculate_first_density_matrix 1 7.0 0.000 0.000 3.006 3.008 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.909 2.948 calculate_dm_sparse 110 9.5 0.001 0.001 2.702 2.736 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.630 2.695 fft3d_ps 1111 14.6 1.090 1.100 2.599 2.622 potential_pw2rs 110 12.3 0.011 0.012 2.406 2.424 grid_collocate_task_list 110 9.6 2.212 2.395 2.212 2.395 multiply_cannon_metrocomm4 14385 15.4 0.049 0.053 0.862 2.327 mp_alltoall_d11v 2046 13.8 1.892 2.311 1.892 2.311 mp_irecv_dv 48980 15.7 0.787 2.194 0.787 2.194 mp_sum_l 6594 12.7 1.504 2.179 1.504 2.179 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.971 1.973 dbcsr_complete_redistribute 325 12.2 0.326 0.351 1.495 1.962 cp_fm_upper_to_full 70 14.2 1.425 1.824 1.425 1.824 cp_fm_cholesky_decompose 22 10.9 1.762 1.780 1.762 1.780 transfer_rs2pw 451 10.6 0.005 0.006 1.558 1.738 acc_transpose_blocks 16440 15.4 0.076 0.078 1.611 1.716 mp_allgather_i34 2055 14.4 0.567 1.710 0.567 1.710 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.692 1.708 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.621 1.633 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.012 1.482 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.371 1.482 transfer_pw2rs 451 13.1 0.005 0.005 1.379 1.392 mp_waitany 17072 13.8 1.174 1.390 1.174 1.390 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.319 1.327 mp_alltoall_z22v 1111 16.6 1.294 1.326 1.294 1.326 rs_gather_matrices 110 12.3 0.137 0.150 0.919 1.300 jit_kernel_multiply 6 16.7 0.319 1.294 0.319 1.294 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.622000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.272727, yerr=7.664451 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_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 736.325632E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.029 67.119 67.121 qs_mol_dyn_low 1 2.0 0.003 0.004 66.871 66.880 qs_forces 11 3.9 0.003 0.003 66.804 66.805 qs_energies 11 4.9 0.002 0.005 63.399 63.403 scf_env_do_scf 11 5.9 0.000 0.001 54.695 54.698 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.902 42.902 velocity_verlet 10 3.0 0.002 0.002 37.718 37.720 dbcsr_multiply_generic 2055 12.4 0.114 0.120 31.139 31.310 qs_scf_new_mos 99 7.5 0.001 0.001 28.170 28.265 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.170 28.264 ot_scf_mini 99 9.5 0.002 0.003 26.470 26.557 multiply_cannon 2055 13.4 0.237 0.256 23.977 25.182 multiply_cannon_loop 2055 14.4 1.406 1.486 22.384 22.953 ot_mini 99 10.5 0.001 0.001 14.928 15.040 multiply_cannon_multrec 24660 15.4 4.109 6.836 13.276 14.488 rebuild_ks_matrix 110 8.3 0.000 0.001 12.063 12.163 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.063 12.163 init_scf_loop 11 6.9 0.001 0.004 11.748 11.748 qs_ot_get_derivative 99 11.5 0.001 0.001 10.749 10.839 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.662 10.751 prepare_preconditioner 11 7.9 0.000 0.000 9.991 10.007 make_preconditioner 11 8.9 0.000 0.001 9.991 10.007 dbcsr_mm_accdrv_process 52304 16.0 7.784 9.126 9.008 9.949 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.246 9.703 qs_ot_get_p 110 10.4 0.001 0.001 6.384 6.524 sum_up_and_integrate 110 10.3 0.001 0.002 6.317 6.330 integrate_v_rspace 110 11.3 0.003 0.003 6.291 6.304 mp_waitall_1 121746 16.5 4.162 6.083 4.162 6.083 init_scf_run 11 5.9 0.000 0.001 5.903 5.903 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.902 5.903 make_m2s 4110 13.4 0.060 0.062 5.440 5.902 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.887 5.900 calculate_rho_elec 110 8.6 0.078 0.081 5.887 5.899 make_images 4110 14.4 0.580 0.701 5.298 5.759 cp_fm_upper_to_full 70 14.2 3.355 4.857 3.355 4.857 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.432 4.447 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.051 4.139 apply_single 110 13.6 0.000 0.000 4.051 4.138 ot_diis_step 99 11.5 0.011 0.011 4.136 4.137 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.856 3.856 dbcsr_complete_redistribute 325 12.2 0.417 0.461 2.675 3.802 pw_transfer 1331 11.6 0.066 0.074 3.722 3.758 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.615 3.655 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.583 3.629 grid_integrate_task_list 110 12.3 3.285 3.449 3.285 3.449 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.393 3.444 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.259 3.269 cp_fm_diag_elpa_base 48 14.0 3.109 3.167 3.257 3.268 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.107 3.224 calculate_dm_sparse 110 9.5 0.001 0.001 3.144 3.181 density_rs2pw 110 9.6 0.004 0.004 3.007 3.134 fft_wrap_pw1pw2_140 451 13.1 0.609 0.630 3.069 3.112 hybrid_alltoall_any 4261 16.3 0.122 0.458 2.284 3.108 make_images_data 4110 15.4 0.049 0.052 2.626 3.101 wfi_extrapolate 11 7.9 0.001 0.001 3.053 3.053 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.378 3.051 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.011 3.013 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.736 2.830 mp_alltoall_i22 605 13.7 1.650 2.825 1.650 2.825 cp_fm_cholesky_invert 11 10.9 2.783 2.792 2.783 2.792 calculate_first_density_matrix 1 7.0 0.000 0.001 2.752 2.756 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.638 2.673 acc_transpose_blocks 24660 15.4 0.113 0.116 2.452 2.599 multiply_cannon_sync_h2d 24660 15.4 2.378 2.551 2.378 2.551 fft3d_ps 1111 14.6 1.085 1.115 2.497 2.514 grid_collocate_task_list 110 9.6 2.262 2.394 2.262 2.394 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.291 2.292 potential_pw2rs 110 12.3 0.013 0.014 2.115 2.120 mp_alltoall_d11v 2046 13.8 1.813 2.060 1.813 2.060 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.819 1.856 cp_fm_cholesky_decompose 22 10.9 1.683 1.731 1.683 1.731 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.617 1.714 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.690 1.709 mp_sum_l 6594 12.7 1.045 1.697 1.045 1.697 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.646 1.656 mp_allgather_i34 2055 14.4 0.674 1.561 0.674 1.561 multiply_cannon_metrocomm4 20550 15.4 0.063 0.065 0.844 1.558 acc_transpose_blocks_sync 73980 16.4 1.392 1.538 1.392 1.538 mp_irecv_dv 62702 16.1 0.737 1.473 0.737 1.473 transfer_rs2pw 451 10.6 0.005 0.006 1.241 1.385 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.121000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=697.727273, yerr=9.516406 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_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 843.165696E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.028 56.045 56.047 qs_mol_dyn_low 1 2.0 0.003 0.003 55.794 55.804 qs_forces 11 3.9 0.003 0.003 55.704 55.705 qs_energies 11 4.9 0.002 0.002 52.048 52.052 scf_env_do_scf 11 5.9 0.000 0.001 43.410 43.410 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.611 35.612 velocity_verlet 10 3.0 0.001 0.001 31.353 31.358 dbcsr_multiply_generic 2055 12.4 0.106 0.110 23.746 23.920 qs_scf_new_mos 99 7.5 0.001 0.001 21.132 21.176 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.132 21.175 ot_scf_mini 99 9.5 0.002 0.002 19.881 19.902 multiply_cannon 2055 13.4 0.235 0.246 18.046 19.447 multiply_cannon_loop 2055 14.4 0.604 0.624 16.761 17.096 rebuild_ks_matrix 110 8.3 0.000 0.000 11.576 11.594 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.576 11.593 ot_mini 99 10.5 0.001 0.001 10.881 10.894 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.288 10.305 multiply_cannon_multrec 8220 15.4 3.200 4.442 7.890 8.793 init_scf_loop 11 6.9 0.000 0.000 7.752 7.753 mp_waitall_1 103326 16.6 5.860 7.329 5.860 7.329 qs_ot_get_derivative 99 11.5 0.001 0.001 7.197 7.216 sum_up_and_integrate 110 10.3 0.001 0.002 6.181 6.193 integrate_v_rspace 110 11.3 0.003 0.004 6.154 6.165 prepare_preconditioner 11 7.9 0.000 0.000 6.113 6.117 make_preconditioner 11 8.9 0.000 0.000 6.113 6.117 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.949 5.961 calculate_rho_elec 110 8.6 0.114 0.114 5.949 5.961 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.699 5.769 dbcsr_mm_accdrv_process 17442 15.9 3.346 4.368 4.552 5.505 init_scf_run 11 5.9 0.000 0.001 5.500 5.500 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.500 5.500 qs_ot_get_p 110 10.4 0.001 0.001 5.029 5.047 make_m2s 4110 13.4 0.038 0.039 4.291 4.571 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 2.980 4.501 make_images 4110 14.4 0.653 0.719 4.162 4.442 pw_transfer 1331 11.6 0.066 0.071 3.845 3.857 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.730 3.776 apply_single 110 13.6 0.000 0.000 3.730 3.776 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.737 3.753 ot_diis_step 99 11.5 0.012 0.012 3.661 3.661 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.600 3.605 grid_integrate_task_list 110 12.3 3.381 3.509 3.381 3.509 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.256 3.256 fft_wrap_pw1pw2_140 451 13.1 0.776 0.788 3.225 3.244 density_rs2pw 110 9.6 0.004 0.004 2.943 3.060 cp_fm_cholesky_invert 11 10.9 2.949 2.953 2.949 2.953 calculate_dm_sparse 110 9.5 0.001 0.001 2.703 2.747 hybrid_alltoall_any 4261 16.3 0.201 0.853 2.284 2.743 wfi_extrapolate 11 7.9 0.001 0.001 2.734 2.734 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.691 2.693 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.676 2.683 cp_fm_diag_elpa_base 48 14.0 2.619 2.642 2.674 2.682 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 2.674 2.675 calculate_first_density_matrix 1 7.0 0.000 0.000 2.670 2.671 make_images_data 4110 15.4 0.041 0.046 2.286 2.646 grid_collocate_task_list 110 9.6 2.362 2.552 2.362 2.552 multiply_cannon_sync_h2d 8220 15.4 2.367 2.463 2.367 2.463 fft3d_ps 1111 14.6 1.135 1.154 2.403 2.419 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.318 2.332 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.061 2.093 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.788 2.001 potential_pw2rs 110 12.3 0.015 0.016 1.972 1.976 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.945 1.957 mp_alltoall_d11v 2046 13.8 1.726 1.880 1.726 1.880 cp_fm_cholesky_decompose 22 10.9 1.714 1.731 1.714 1.731 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.660 1.672 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.616 1.618 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.500 1.612 mp_allgather_i34 2055 14.4 0.456 1.556 0.456 1.556 dbcsr_complete_redistribute 325 12.2 0.558 0.590 1.458 1.555 acc_transpose_blocks 8220 15.4 0.039 0.040 1.482 1.508 qs_create_task_list 11 7.9 0.000 0.000 1.212 1.309 generate_qs_task_list 11 8.9 0.374 0.441 1.211 1.308 transfer_rs2pw 451 10.6 0.005 0.005 1.143 1.286 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.228 1.257 jit_kernel_multiply 7 15.6 0.884 1.242 0.884 1.242 mp_waitany 9240 13.8 1.054 1.179 1.054 1.179 multiply_cannon_metrocomm1 8220 15.4 0.022 0.022 0.754 1.157 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.047000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=797.545455, yerr=9.394037 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_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 1.391440E+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 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.091 88.524 88.576 qs_mol_dyn_low 1 2.0 0.003 0.003 88.199 88.208 qs_forces 11 3.9 0.003 0.003 88.084 88.085 qs_energies 11 4.9 0.001 0.001 83.952 83.953 scf_env_do_scf 11 5.9 0.000 0.001 73.624 73.624 velocity_verlet 10 3.0 0.002 0.002 55.933 55.945 scf_env_do_scf_inner_loop 99 6.5 0.004 0.007 45.248 45.249 dbcsr_multiply_generic 2055 12.4 0.120 0.126 30.311 30.391 init_scf_loop 11 6.9 0.000 0.000 28.301 28.302 qs_scf_new_mos 99 7.5 0.001 0.001 27.773 27.829 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.772 27.829 prepare_preconditioner 11 7.9 0.000 0.000 26.299 26.307 make_preconditioner 11 8.9 0.000 0.000 26.299 26.307 ot_scf_mini 99 9.5 0.002 0.002 25.945 25.989 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.643 25.751 multiply_cannon 2055 13.4 0.330 0.356 22.930 23.605 multiply_cannon_loop 2055 14.4 0.823 0.838 21.124 21.707 cp_fm_upper_to_full 70 14.2 12.618 17.968 12.618 17.968 ot_mini 99 10.5 0.001 0.001 14.594 14.640 rebuild_ks_matrix 110 8.3 0.001 0.001 13.679 13.723 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.678 13.722 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.421 12.461 dbcsr_complete_redistribute 325 12.2 1.020 1.037 7.287 10.361 multiply_cannon_multrec 8220 15.4 4.048 4.216 9.908 10.028 qs_ot_get_derivative 99 11.5 0.001 0.001 9.944 9.988 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.221 9.290 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.640 8.688 mp_waitall_1 84994 16.7 7.758 8.677 7.758 8.677 mp_alltoall_i22 605 13.7 5.263 8.348 5.263 8.348 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.193 7.234 calculate_rho_elec 110 8.6 0.224 0.224 7.193 7.233 sum_up_and_integrate 110 10.3 0.002 0.002 6.847 6.860 integrate_v_rspace 110 11.3 0.004 0.004 6.819 6.832 qs_ot_get_p 110 10.4 0.001 0.001 6.013 6.063 init_scf_run 11 5.9 0.000 0.001 6.049 6.049 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.048 6.049 dbcsr_mm_accdrv_process 11614 15.7 3.814 4.156 5.713 5.941 make_m2s 4110 13.4 0.043 0.044 5.321 5.805 make_images 4110 14.4 0.886 0.941 5.131 5.614 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.043 5.551 apply_single 110 13.6 0.000 0.000 5.043 5.551 cp_fm_cholesky_invert 11 10.9 5.432 5.436 5.432 5.436 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 5.132 5.393 pw_transfer 1331 11.6 0.075 0.075 4.947 4.953 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.830 4.836 ot_diis_step 99 11.5 0.015 0.015 4.628 4.628 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.382 4.389 fft_wrap_pw1pw2_140 451 13.1 1.278 1.282 4.195 4.202 grid_integrate_task_list 110 12.3 3.723 3.875 3.723 3.875 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.862 3.862 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.096 3.815 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.662 3.662 density_rs2pw 110 9.6 0.004 0.004 3.608 3.660 hybrid_alltoall_any 4261 16.3 0.264 0.564 2.820 3.442 make_images_data 4110 15.4 0.045 0.048 2.752 3.379 wfi_extrapolate 11 7.9 0.001 0.001 3.357 3.357 calculate_dm_sparse 110 9.5 0.001 0.001 3.283 3.311 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.242 3.243 cp_fm_diag_elpa_base 48 14.0 2.704 2.897 3.240 3.240 multiply_cannon_sync_h2d 8220 15.4 3.130 3.163 3.130 3.163 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.156 3.159 fft3d_ps 1111 14.6 1.293 1.298 2.885 2.894 grid_collocate_task_list 110 9.6 2.705 2.789 2.705 2.789 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.706 2.733 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.628 2.640 calculate_first_density_matrix 1 7.0 0.000 0.000 2.574 2.575 mp_alltoall_d11v 2046 13.8 2.165 2.263 2.165 2.263 potential_pw2rs 110 12.3 0.022 0.022 2.250 2.251 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.176 2.233 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.119 2.190 cp_fm_cholesky_decompose 22 10.9 2.071 2.090 2.071 2.090 acc_transpose_blocks 8220 15.4 0.041 0.041 1.756 2.044 jit_kernel_multiply 10 15.2 1.699 2.014 1.699 2.014 qs_create_task_list 11 7.9 0.001 0.001 1.883 1.928 generate_qs_task_list 11 8.9 0.731 0.783 1.882 1.927 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.858 1.863 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.814 1.845 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.758 1.772 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.576000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1241.181818, yerr=56.253999 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 1.103326E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 12044928 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.1 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 633.532416E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881156E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148350408 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57450. MP_Allreduce 11139 794. MP_Sync 87 MP_Alltoall 2242 3013554. MP_SendRecv 24510 18752. MP_ISendRecv 24510 18752. MP_Wait 42802 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.031 209.884 209.885 qs_mol_dyn_low 1 2.0 0.003 0.003 209.463 209.477 qs_forces 11 3.9 0.004 0.005 209.362 209.365 qs_energies 11 4.9 0.001 0.002 203.789 203.801 scf_env_do_scf 11 5.9 0.001 0.001 186.318 186.323 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 165.915 165.918 dbcsr_multiply_generic 2527 12.6 0.182 0.185 128.053 128.587 qs_scf_new_mos 118 7.6 0.001 0.001 126.259 126.451 qs_scf_loop_do_ot 118 8.6 0.001 0.001 126.258 126.450 velocity_verlet 10 3.0 0.001 0.002 124.946 124.947 ot_scf_mini 118 9.6 0.003 0.003 119.504 119.699 multiply_cannon 2527 13.6 0.241 0.249 103.562 104.840 multiply_cannon_loop 2527 14.6 2.422 2.476 101.267 102.739 ot_mini 118 10.6 0.001 0.001 67.617 67.846 multiply_cannon_multrec 60648 15.6 31.914 33.640 42.405 44.559 qs_ot_get_derivative 118 11.6 0.001 0.001 42.779 42.954 rebuild_ks_matrix 129 8.3 0.001 0.001 33.664 33.976 qs_ks_build_kohn_sham_matrix 129 9.3 0.015 0.016 33.663 33.975 mp_waitall_1 269254 16.5 29.822 32.495 29.822 32.495 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.261 30.568 qs_ot_get_p 129 10.4 0.001 0.001 29.608 29.793 multiply_cannon_sync_h2d 60648 15.6 26.305 28.197 26.305 28.197 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.352 25.190 apply_single 129 13.6 0.001 0.001 24.351 25.190 ot_diis_step 118 11.6 0.008 0.008 24.576 24.577 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.894 22.958 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.805 20.967 init_scf_loop 11 6.9 0.000 0.000 20.329 20.330 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.195 20.196 multiply_cannon_metrocomm3 60648 15.6 0.117 0.122 16.324 18.244 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.209 17.249 cp_fm_diag_elpa_base 83 14.4 17.139 17.181 17.205 17.245 prepare_preconditioner 11 7.9 0.000 0.000 15.783 15.814 make_preconditioner 11 8.9 0.000 0.000 15.783 15.814 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.024 15.212 make_m2s 5054 13.6 0.106 0.116 13.878 14.175 sum_up_and_integrate 129 10.3 0.002 0.004 14.073 14.086 integrate_v_rspace 129 11.3 0.003 0.004 14.014 14.031 make_images 5054 14.6 0.402 0.421 13.693 14.002 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.276 13.437 calculate_rho_elec 129 8.7 0.045 0.064 13.276 13.436 init_scf_run 11 5.9 0.000 0.001 13.357 13.358 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.357 13.358 dbcsr_mm_accdrv_process 125468 16.2 4.804 4.953 10.060 11.031 mp_sum_l 8010 12.9 9.343 10.648 9.343 10.648 calculate_dm_sparse 129 9.5 0.001 0.001 9.409 9.545 wfi_extrapolate 11 7.9 0.001 0.001 9.052 9.052 cp_fm_cholesky_invert 11 10.9 8.986 8.993 8.986 8.993 multiply_cannon_metrocomm1 60648 15.6 0.093 0.097 6.988 8.878 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.337 8.464 qs_ot_get_orbitals 118 10.6 0.001 0.001 8.081 8.147 pw_transfer 1559 11.6 0.075 0.094 7.654 7.841 make_images_data 5054 15.6 0.067 0.072 6.800 7.732 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 7.450 7.641 grid_integrate_task_list 129 12.3 7.125 7.547 7.125 7.547 density_rs2pw 129 9.7 0.006 0.007 6.847 7.422 hybrid_alltoall_any 5240 16.5 0.298 2.286 5.954 7.203 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.898 6.911 fft_wrap_pw1pw2_140 527 13.2 1.134 1.179 6.518 6.706 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.052 6.134 fft3d_ps 1301 14.7 2.211 2.855 5.442 5.762 mp_alltoall_d11v 2423 14.1 4.333 5.533 4.333 5.533 grid_collocate_task_list 129 9.7 4.882 5.315 4.882 5.315 cp_fm_cholesky_decompose 22 10.9 4.635 4.652 4.635 4.652 potential_pw2rs 129 12.3 0.009 0.010 4.514 4.541 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=209.885000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.727273, yerr=6.620779 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 844.902400E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57958. MP_Allreduce 11128 958. MP_Sync 87 MP_Alltoall 1969 4687307. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.030 190.507 190.508 qs_mol_dyn_low 1 2.0 0.003 0.003 189.617 189.630 qs_forces 11 3.9 0.004 0.004 189.532 189.533 qs_energies 11 4.9 0.001 0.002 182.784 182.795 scf_env_do_scf 11 5.9 0.001 0.001 166.117 166.128 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 133.184 133.187 velocity_verlet 10 3.0 0.001 0.002 119.167 119.169 dbcsr_multiply_generic 2507 12.6 0.190 0.196 97.703 99.010 qs_scf_new_mos 117 7.6 0.001 0.001 94.645 95.149 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.644 95.148 ot_scf_mini 117 9.6 0.004 0.004 89.829 90.420 multiply_cannon 2507 13.6 0.502 0.556 77.605 82.204 multiply_cannon_loop 2507 14.6 1.573 1.637 74.216 76.821 ot_mini 117 10.6 0.001 0.001 50.187 50.703 mp_waitall_1 214728 16.6 23.927 37.429 23.927 37.429 multiply_cannon_multrec 30084 15.6 20.944 25.685 32.042 37.343 rebuild_ks_matrix 128 8.3 0.001 0.001 32.242 32.878 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.241 32.877 init_scf_loop 11 6.9 0.000 0.000 32.838 32.839 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.999 29.591 qs_ot_get_derivative 117 11.6 0.001 0.002 28.236 28.815 prepare_preconditioner 11 7.9 0.000 0.000 28.534 28.605 make_preconditioner 11 8.9 0.000 0.000 28.534 28.605 multiply_cannon_metrocomm3 30084 15.6 0.097 0.104 15.569 28.197 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.225 27.769 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.051 23.316 apply_single 128 13.6 0.001 0.001 22.051 23.315 qs_ot_get_p 128 10.4 0.001 0.001 21.701 22.324 ot_diis_step 117 11.6 0.014 0.015 21.764 21.766 multiply_cannon_sync_h2d 30084 15.6 17.930 19.942 17.930 19.942 qs_ot_p2m_diag 83 11.4 0.189 0.217 16.778 16.812 cp_fm_cholesky_invert 11 10.9 16.620 16.632 16.620 16.632 make_m2s 5014 13.6 0.090 0.098 14.177 15.870 make_images 5014 14.6 1.179 1.374 13.963 15.656 cp_dbcsr_syevd 83 12.4 0.006 0.006 15.557 15.558 sum_up_and_integrate 128 10.3 0.002 0.004 14.015 14.047 integrate_v_rspace 128 11.3 0.003 0.004 13.955 13.991 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.395 13.430 calculate_rho_elec 128 8.7 0.087 0.104 13.395 13.429 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.349 12.390 cp_fm_diag_elpa_base 83 14.4 12.094 12.186 12.343 12.380 init_scf_run 11 5.9 0.000 0.001 11.869 11.871 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.869 11.871 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.175 11.581 dbcsr_mm_accdrv_process 62242 16.2 5.500 6.074 10.552 11.201 multiply_cannon_metrocomm4 27577 15.6 0.107 0.119 3.882 10.787 mp_irecv_dv 69486 16.3 3.675 10.390 3.675 10.390 make_images_data 5014 15.6 0.066 0.076 8.304 10.206 hybrid_alltoall_any 5200 16.5 0.350 1.541 7.062 9.772 pw_transfer 1547 11.6 0.085 0.097 8.556 8.620 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.333 8.400 wfi_extrapolate 11 7.9 0.001 0.001 8.325 8.325 grid_integrate_task_list 128 12.3 7.198 7.537 7.198 7.537 fft_wrap_pw1pw2_140 523 13.2 1.214 1.237 7.377 7.467 density_rs2pw 128 9.7 0.006 0.007 7.080 7.441 cp_fm_cholesky_decompose 22 10.9 6.984 7.063 6.984 7.063 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.333 7.057 calculate_dm_sparse 128 9.5 0.001 0.001 6.488 6.638 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.531 6.541 mp_sum_l 7950 12.9 4.117 6.237 4.117 6.237 fft3d_ps 1291 14.7 2.838 3.010 5.892 5.942 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.643 5.807 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.376 5.448 grid_collocate_task_list 128 9.7 5.007 5.305 5.007 5.305 mp_allgather_i34 2507 14.6 1.902 5.081 1.902 5.081 mp_alltoall_d11v 2415 14.1 4.138 4.586 4.138 4.586 potential_pw2rs 128 12.3 0.016 0.019 4.509 4.531 mp_sum_d 4479 12.1 2.601 4.017 2.601 4.017 dbcsr_complete_redistribute 395 12.7 0.782 0.888 3.171 4.016 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.508000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=804.818182, yerr=2.516064 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 935.387136E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172558800 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58307. MP_Allreduce 11065 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.054 173.966 173.983 qs_mol_dyn_low 1 2.0 0.003 0.003 173.358 173.542 qs_forces 11 3.9 0.004 0.004 173.263 173.266 qs_energies 11 4.9 0.001 0.002 166.709 166.719 scf_env_do_scf 11 5.9 0.001 0.001 150.624 150.625 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.950 115.951 velocity_verlet 10 3.0 0.001 0.002 110.623 110.625 dbcsr_multiply_generic 2485 12.5 0.180 0.185 81.780 82.924 qs_scf_new_mos 116 7.6 0.001 0.001 79.456 79.759 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.456 79.758 ot_scf_mini 116 9.6 0.003 0.004 75.235 75.608 multiply_cannon 2485 13.5 0.504 0.524 62.375 67.029 multiply_cannon_loop 2485 14.5 1.120 1.178 59.331 61.916 ot_mini 116 10.6 0.001 0.001 42.012 42.373 init_scf_loop 11 6.9 0.000 0.000 34.574 34.575 mp_waitall_1 169034 16.6 24.096 32.818 24.096 32.818 prepare_preconditioner 11 7.9 0.000 0.000 30.609 30.651 make_preconditioner 11 8.9 0.000 0.000 30.609 30.651 rebuild_ks_matrix 127 8.3 0.001 0.001 29.790 30.266 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 29.790 30.266 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.267 29.665 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.807 27.242 multiply_cannon_multrec 19880 15.5 13.108 16.271 23.132 26.292 multiply_cannon_metrocomm3 19880 15.5 0.062 0.065 14.888 23.737 qs_ot_get_derivative 116 11.6 0.001 0.002 22.730 23.104 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.451 20.405 apply_single 127 13.6 0.001 0.001 19.451 20.405 qs_ot_get_p 127 10.4 0.001 0.001 18.741 19.178 ot_diis_step 116 11.6 0.018 0.018 19.175 19.176 make_m2s 4970 13.5 0.079 0.085 14.452 15.716 make_images 4970 14.5 1.148 1.228 14.219 15.475 multiply_cannon_sync_h2d 19880 15.5 13.536 14.835 13.536 14.835 qs_ot_p2m_diag 82 11.4 0.262 0.269 14.499 14.508 cp_fm_cholesky_invert 11 10.9 14.225 14.234 14.225 14.234 sum_up_and_integrate 127 10.3 0.002 0.003 13.704 13.726 integrate_v_rspace 127 11.3 0.003 0.004 13.645 13.669 cp_dbcsr_syevd 82 12.4 0.005 0.006 13.462 13.463 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.354 13.384 calculate_rho_elec 127 8.7 0.129 0.143 13.354 13.384 init_scf_run 11 5.9 0.000 0.001 11.099 11.099 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.099 11.099 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.313 10.331 cp_fm_diag_elpa_base 82 14.4 9.905 10.073 10.310 10.328 make_images_data 4970 15.5 0.062 0.072 8.654 10.319 hybrid_alltoall_any 5155 16.4 0.444 2.005 7.603 9.882 dbcsr_mm_accdrv_process 41158 16.2 6.285 6.600 9.486 9.614 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.926 9.176 multiply_cannon_metrocomm4 17395 15.5 0.067 0.076 3.392 9.113 mp_irecv_dv 49801 16.2 3.263 8.860 3.263 8.860 pw_transfer 1535 11.6 0.085 0.102 8.535 8.640 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 8.312 8.425 grid_integrate_task_list 127 12.3 7.267 7.687 7.267 7.687 fft_wrap_pw1pw2_140 519 13.2 1.291 1.320 7.399 7.521 wfi_extrapolate 11 7.9 0.001 0.001 7.426 7.426 cp_fm_upper_to_full 104 14.8 5.790 7.324 5.790 7.324 cp_fm_cholesky_decompose 22 10.9 7.181 7.204 7.181 7.204 density_rs2pw 127 9.7 0.006 0.006 6.768 7.047 dbcsr_complete_redistribute 393 12.7 1.174 1.206 4.539 6.282 calculate_dm_sparse 127 9.5 0.001 0.001 6.141 6.246 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.814 5.821 fft3d_ps 1281 14.7 2.738 2.955 5.710 5.788 grid_collocate_task_list 127 9.7 5.120 5.407 5.120 5.407 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.587 5.272 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.390 5.135 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.932 5.064 mp_alltoall_d11v 2401 14.1 4.166 4.822 4.166 4.822 mp_allgather_i34 2485 14.5 1.544 4.699 1.544 4.699 mp_sum_l 7884 12.9 3.131 4.588 3.131 4.588 potential_pw2rs 127 12.3 0.021 0.023 4.241 4.250 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.323 4.012 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.867 3.885 mp_alltoall_i22 712 14.1 1.949 3.849 1.949 3.849 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.715 3.715 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.476 3.516 calculate_first_density_matrix 1 7.0 0.000 0.000 3.488 3.490 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.983000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=886.272727, yerr=8.136237 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 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 1.142972E+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 669922014800 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57626. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509628. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.061 188.857 188.865 qs_mol_dyn_low 1 2.0 0.003 0.003 188.227 188.488 qs_forces 11 3.9 0.004 0.004 188.108 188.113 qs_energies 11 4.9 0.021 0.022 180.988 180.999 scf_env_do_scf 11 5.9 0.001 0.001 164.258 164.272 velocity_verlet 10 3.0 0.002 0.002 125.337 125.341 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 118.090 118.091 dbcsr_multiply_generic 2529 12.6 0.190 0.197 81.458 82.217 qs_scf_new_mos 118 7.6 0.001 0.001 81.753 82.073 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.752 82.072 ot_scf_mini 118 9.6 0.003 0.003 77.117 77.465 multiply_cannon 2529 13.6 0.555 0.588 56.455 59.250 multiply_cannon_loop 2529 14.6 1.852 1.945 52.776 54.686 init_scf_loop 11 6.9 0.000 0.000 46.047 46.048 ot_mini 118 10.6 0.001 0.001 43.636 43.976 prepare_preconditioner 11 7.9 0.000 0.000 41.986 42.009 make_preconditioner 11 8.9 0.000 0.000 41.986 42.009 make_full_inverse_cholesky 11 9.9 0.012 0.023 35.659 40.584 multiply_cannon_multrec 30348 15.6 13.652 19.139 26.717 31.830 rebuild_ks_matrix 129 8.3 0.001 0.001 29.357 29.677 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 29.357 29.676 mp_waitall_1 149172 16.7 17.221 27.005 17.221 27.005 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.491 26.780 qs_ot_get_derivative 118 11.6 0.002 0.002 23.802 24.156 make_m2s 5058 13.6 0.097 0.102 20.518 21.453 make_images 5058 14.6 1.990 2.316 20.204 21.143 qs_ot_get_p 129 10.4 0.001 0.001 19.510 19.887 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.156 19.749 apply_single 129 13.6 0.001 0.001 19.155 19.748 ot_diis_step 118 11.6 0.018 0.019 19.701 19.703 cp_fm_upper_to_full 106 14.8 11.381 16.712 11.381 16.712 cp_fm_cholesky_invert 11 10.9 16.031 16.040 16.031 16.040 qs_ot_p2m_diag 84 11.4 0.347 0.395 15.266 15.321 multiply_cannon_metrocomm3 30348 15.6 0.050 0.054 6.335 14.946 sum_up_and_integrate 129 10.3 0.002 0.003 13.994 14.023 integrate_v_rspace 129 11.3 0.003 0.004 13.933 13.964 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.848 13.848 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.727 13.779 calculate_rho_elec 129 8.7 0.174 0.190 13.727 13.778 dbcsr_mm_accdrv_process 62780 16.2 8.463 9.274 12.636 13.113 make_images_data 5058 15.6 0.066 0.074 10.990 12.849 dbcsr_complete_redistribute 397 12.7 1.513 1.634 8.900 12.584 hybrid_alltoall_any 5245 16.5 0.534 2.219 9.851 11.725 multiply_cannon_sync_h2d 30348 15.6 10.551 11.356 10.551 11.356 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.549 11.252 init_scf_run 11 5.9 0.000 0.001 10.806 10.807 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.805 10.807 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.706 10.719 cp_fm_diag_elpa_base 84 14.4 9.747 10.080 10.699 10.711 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.919 10.174 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.304 9.932 mp_alltoall_i22 720 14.1 5.499 9.145 5.499 9.145 pw_transfer 1559 11.6 0.086 0.101 8.987 9.068 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.760 8.845 grid_integrate_task_list 129 12.3 7.583 7.910 7.583 7.910 fft_wrap_pw1pw2_140 527 13.2 1.455 1.486 7.809 7.910 wfi_extrapolate 11 7.9 0.001 0.001 7.740 7.740 cp_fm_cholesky_decompose 22 10.9 7.549 7.653 7.549 7.653 multiply_cannon_metrocomm4 25290 15.6 0.086 0.097 2.913 7.184 density_rs2pw 129 9.7 0.006 0.006 6.766 7.064 mp_irecv_dv 76751 16.2 2.752 6.902 2.752 6.902 calculate_dm_sparse 129 9.5 0.001 0.001 6.330 6.440 fft3d_ps 1301 14.7 2.870 2.941 5.920 5.972 grid_collocate_task_list 129 9.7 5.321 5.630 5.321 5.630 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.539 5.608 mp_alltoall_d11v 2429 14.1 4.737 5.167 4.737 5.167 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.834 4.933 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.428 4.552 qs_energies_init_hamiltonians 11 5.9 0.003 0.006 4.418 4.419 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.275 4.342 potential_pw2rs 129 12.3 0.023 0.023 4.231 4.246 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.865000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1078.272727, yerr=20.073007 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 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 1.536786E+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 716108580288 ------------------------------------------------------------------------------- - - - 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 4003 58187. MP_Allreduce 11085 1167. MP_Sync 86 MP_Alltoall 1700 18828189. 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 0.013 0.034 171.733 171.735 qs_mol_dyn_low 1 2.0 0.003 0.003 171.325 171.339 qs_forces 11 3.9 0.004 0.005 171.173 171.176 qs_energies 11 4.9 0.002 0.002 163.734 163.743 scf_env_do_scf 11 5.9 0.001 0.001 146.303 146.311 velocity_verlet 10 3.0 0.001 0.001 111.610 111.618 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 110.375 110.376 qs_scf_new_mos 116 7.6 0.001 0.001 73.842 73.924 qs_scf_loop_do_ot 116 8.6 0.001 0.001 73.841 73.923 dbcsr_multiply_generic 2485 12.5 0.183 0.195 71.904 72.150 ot_scf_mini 116 9.6 0.003 0.004 69.427 69.482 multiply_cannon 2485 13.5 0.555 0.581 53.099 56.621 multiply_cannon_loop 2485 14.5 0.802 0.837 50.080 50.780 ot_mini 116 10.6 0.001 0.001 37.355 37.414 init_scf_loop 11 6.9 0.000 0.000 35.774 35.776 prepare_preconditioner 11 7.9 0.000 0.000 31.906 31.922 make_preconditioner 11 8.9 0.000 0.000 31.906 31.922 mp_waitall_1 124680 16.7 24.469 30.743 24.469 30.743 make_full_inverse_cholesky 11 9.9 0.014 0.025 29.723 29.974 rebuild_ks_matrix 127 8.3 0.001 0.001 28.892 29.012 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 28.891 29.011 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.318 26.432 multiply_cannon_multrec 9940 15.5 10.322 14.872 18.144 21.332 qs_ot_get_derivative 116 11.6 0.001 0.002 20.546 20.605 qs_ot_get_p 127 10.4 0.001 0.001 19.426 19.483 multiply_cannon_metrocomm3 9940 15.5 0.026 0.027 12.341 19.433 cp_fm_cholesky_invert 11 10.9 18.223 18.229 18.223 18.229 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 16.878 17.042 apply_single 127 13.6 0.001 0.001 16.878 17.041 ot_diis_step 116 11.6 0.020 0.020 16.741 16.742 make_m2s 4970 13.5 0.065 0.069 14.892 15.768 qs_ot_p2m_diag 82 11.4 0.490 0.495 15.529 15.544 make_images 4970 14.5 2.164 2.588 14.586 15.465 cp_dbcsr_syevd 82 12.4 0.005 0.006 14.358 14.359 sum_up_and_integrate 127 10.3 0.002 0.002 14.036 14.085 integrate_v_rspace 127 11.3 0.004 0.004 13.976 14.027 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.967 13.986 calculate_rho_elec 127 8.7 0.254 0.265 13.966 13.986 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.391 11.402 cp_fm_diag_elpa_base 82 14.4 11.147 11.228 11.384 11.396 multiply_cannon_sync_h2d 9940 15.5 10.789 11.204 10.789 11.204 init_scf_run 11 5.9 0.000 0.001 10.627 10.627 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.626 10.627 hybrid_alltoall_any 5155 16.4 0.837 3.762 8.377 9.899 make_images_data 4970 15.5 0.055 0.065 8.546 9.856 pw_transfer 1535 11.6 0.085 0.094 9.316 9.345 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.095 9.130 dbcsr_mm_accdrv_process 20590 16.1 3.332 4.673 7.475 8.202 grid_integrate_task_list 127 12.3 7.712 8.076 7.712 8.076 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.026 8.064 fft_wrap_pw1pw2_140 519 13.2 1.785 1.827 8.013 8.052 cp_fm_cholesky_decompose 22 10.9 7.927 8.015 7.927 8.015 wfi_extrapolate 11 7.9 0.001 0.001 7.392 7.393 multiply_cannon_metrocomm1 9940 15.5 0.029 0.030 4.298 7.244 density_rs2pw 127 9.7 0.005 0.006 6.745 7.127 calculate_dm_sparse 127 9.5 0.001 0.001 6.222 6.276 fft3d_ps 1281 14.7 2.747 2.812 5.856 5.886 grid_collocate_task_list 127 9.7 5.519 5.833 5.519 5.833 dbcsr_complete_redistribute 393 12.7 2.117 2.189 5.384 5.765 mp_alltoall_d11v 2401 14.1 4.725 5.356 4.725 5.356 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.245 5.255 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.232 5.234 mp_allgather_i34 2485 14.5 1.214 4.667 1.214 4.667 potential_pw2rs 127 12.3 0.026 0.028 4.274 4.283 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.138 4.190 multiply_cannon_metrocomm4 7455 15.5 0.026 0.029 1.724 3.980 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.611 3.921 mp_irecv_dv 28618 15.9 1.683 3.906 1.683 3.906 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.475 3.784 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.619 3.696 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.536 3.620 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.560 3.588 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.469 3.478 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=171.735000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1443.181818, yerr=22.818725 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 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 11.786061E+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 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.154297E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58558. MP_Allreduce 11184 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.445 294.548 294.898 qs_mol_dyn_low 1 2.0 0.003 0.003 293.848 293.870 qs_forces 11 3.9 0.004 0.004 293.749 293.752 qs_energies 11 4.9 0.001 0.002 284.913 284.918 scf_env_do_scf 11 5.9 0.001 0.001 261.808 261.817 velocity_verlet 10 3.0 0.002 0.002 212.265 212.273 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.286 136.289 init_scf_loop 11 6.9 0.000 0.000 125.243 125.250 prepare_preconditioner 11 7.9 0.000 0.000 120.449 120.468 make_preconditioner 11 8.9 0.000 0.000 120.449 120.468 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.979 117.604 qs_scf_new_mos 118 7.6 0.001 0.001 92.564 92.615 qs_scf_loop_do_ot 118 8.6 0.001 0.001 92.563 92.614 ot_scf_mini 118 9.6 0.004 0.004 87.608 87.640 dbcsr_multiply_generic 2529 12.6 0.215 0.227 83.802 84.199 cp_fm_upper_to_full 106 14.8 53.306 76.829 53.306 76.829 multiply_cannon 2529 13.6 0.682 0.714 60.683 61.674 multiply_cannon_loop 2529 14.6 1.050 1.068 56.624 57.861 ot_mini 118 10.6 0.001 0.001 45.097 45.132 dbcsr_complete_redistribute 397 12.7 4.005 4.064 30.489 43.849 copy_fm_to_dbcsr 210 11.7 0.001 0.001 26.942 40.227 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 24.427 37.589 mp_alltoall_i22 720 14.1 22.202 35.314 22.202 35.314 rebuild_ks_matrix 129 8.3 0.001 0.001 34.135 34.167 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 34.135 34.166 cp_fm_cholesky_invert 11 10.9 33.556 33.562 33.556 33.562 mp_waitall_1 104580 16.8 28.380 32.802 28.380 32.802 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.694 31.724 qs_ot_get_p 129 10.4 0.001 0.001 27.244 27.285 qs_ot_get_derivative 118 11.6 0.002 0.002 25.237 25.270 qs_ot_p2m_diag 84 11.4 0.890 0.896 23.065 23.097 cp_dbcsr_syevd 84 12.4 0.006 0.006 21.220 21.221 multiply_cannon_metrocomm3 10116 15.6 0.025 0.027 19.461 20.667 make_m2s 5058 13.6 0.076 0.079 18.387 19.844 ot_diis_step 118 11.6 0.022 0.022 19.826 19.826 make_images 5058 14.6 3.099 3.281 17.902 19.359 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.118 19.230 apply_single 129 13.6 0.001 0.001 19.117 19.230 multiply_cannon_multrec 10116 15.6 10.449 12.344 19.111 19.185 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.933 17.934 cp_fm_diag_elpa_base 84 14.4 13.463 15.115 17.929 17.929 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.621 16.630 calculate_rho_elec 129 8.7 0.484 0.485 16.620 16.629 sum_up_and_integrate 129 10.3 0.002 0.002 16.142 16.237 integrate_v_rspace 129 11.3 0.005 0.005 16.079 16.174 multiply_cannon_sync_h2d 10116 15.6 14.425 14.435 14.425 14.435 init_scf_run 11 5.9 0.000 0.001 12.837 12.838 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.837 12.838 hybrid_alltoall_any 5245 16.5 1.325 3.093 10.357 12.478 make_images_data 5058 15.6 0.064 0.070 10.198 12.354 pw_transfer 1559 11.6 0.097 0.099 11.459 11.467 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 11.218 11.227 dbcsr_mm_accdrv_process 20934 16.1 4.416 6.326 8.411 10.339 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.987 10.018 fft_wrap_pw1pw2_140 527 13.2 3.063 3.096 9.967 9.975 cp_fm_cholesky_decompose 22 10.9 9.016 9.038 9.016 9.038 wfi_extrapolate 11 7.9 0.001 0.001 8.994 8.994 grid_integrate_task_list 129 12.3 8.616 8.807 8.616 8.807 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.100 8.101 density_rs2pw 129 9.7 0.005 0.005 7.576 7.627 calculate_dm_sparse 129 9.5 0.001 0.001 7.167 7.224 mp_alltoall_d11v 2429 14.1 6.895 7.015 6.895 7.015 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.779 6.803 grid_collocate_task_list 129 9.7 6.468 6.542 6.468 6.542 fft3d_ps 1301 14.7 2.870 2.880 6.435 6.492 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.215 6.309 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=294.898000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2878.818182, yerr=145.325227 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_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 1.259516E+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 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.106 85.300 85.301 qs_energies 1 2.0 0.000 0.000 84.694 84.718 ls_scf 1 3.0 0.002 0.015 83.711 83.723 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.698 72.833 multiply_cannon 111 7.7 0.017 0.020 55.929 57.153 multiply_cannon_loop 111 8.7 0.225 0.240 52.534 53.898 ls_scf_main 1 4.0 0.000 0.002 52.423 52.424 density_matrix_trs4 2 5.0 0.002 0.003 46.916 46.999 ls_scf_init_scf 1 4.0 0.000 0.000 28.271 28.273 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.068 27.120 mp_waitall_1 11031 10.9 22.504 26.154 22.504 26.154 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.970 24.985 multiply_cannon_multrec 2664 9.7 8.166 8.880 15.669 17.414 multiply_cannon_sync_h2d 2664 9.7 13.491 14.896 13.491 14.896 make_m2s 222 7.7 0.009 0.011 13.124 13.621 make_images 222 8.7 0.098 0.109 13.102 13.601 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.814 12.591 make_images_data 222 9.7 0.004 0.005 7.658 8.255 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.364 8.185 dbcsr_mm_accdrv_process 4760 10.4 0.599 0.717 7.120 8.117 hybrid_alltoall_any 227 10.6 0.216 1.843 6.580 7.929 dbcsr_mm_accdrv_process_sort 4760 11.4 6.322 7.233 6.322 7.233 calculate_norms 4752 9.8 5.489 6.199 5.489 6.199 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.009 5.111 mp_sum_l 887 5.1 3.125 4.452 3.125 4.452 make_images_sizes 222 9.7 0.000 0.000 0.763 3.455 mp_alltoall_i44 222 10.7 0.763 3.454 0.763 3.454 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.329 3.444 arnoldi_extremal 4 6.8 0.000 0.000 3.228 3.249 arnoldi_normal_ev 4 7.8 0.001 0.003 3.228 3.249 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.057 3.248 mp_irecv_dv 6231 10.9 2.040 3.219 2.040 3.219 build_subspace 16 8.4 0.009 0.012 3.133 3.135 ls_scf_post 1 4.0 0.000 0.001 3.016 3.022 ls_scf_store_result 1 5.0 0.000 0.000 2.839 2.879 dbcsr_special_finalize 555 9.7 0.005 0.006 2.351 2.856 dbcsr_merge_single_wm 555 10.7 0.460 0.585 2.342 2.846 make_images_pack 222 9.7 2.206 2.640 2.208 2.641 dbcsr_sort_data 658 11.4 2.140 2.624 2.140 2.624 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.331 2.572 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.465 2.071 2.467 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.281 2.344 buffer_matrices_ensure_size 222 8.7 1.755 2.083 1.755 2.083 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.808 1.812 rebuild_ks_matrix 3 7.3 0.000 0.000 1.799 1.802 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.030 1.799 1.802 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.301000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_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 2.139087E+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 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.060 91.150 91.152 qs_energies 1 2.0 0.003 0.025 90.334 90.388 ls_scf 1 3.0 0.002 0.014 88.998 89.030 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.999 75.321 multiply_cannon 111 7.7 0.028 0.043 52.913 58.103 ls_scf_main 1 4.0 0.001 0.027 54.945 54.950 multiply_cannon_loop 111 8.7 0.136 0.147 50.062 53.703 density_matrix_trs4 2 5.0 0.002 0.003 49.302 49.514 mp_waitall_1 9105 10.9 20.874 30.632 20.874 30.632 ls_scf_init_scf 1 4.0 0.000 0.001 30.154 30.156 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.923 29.023 multiply_cannon_multrec 1332 9.7 13.341 16.691 22.760 27.224 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.568 26.577 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.555 20.854 make_m2s 222 7.7 0.006 0.007 15.200 15.822 make_images 222 8.7 1.367 1.707 15.169 15.792 dbcsr_mm_accdrv_process 4041 10.4 0.348 0.546 9.012 10.638 dbcsr_mm_accdrv_process_sort 4041 11.4 8.528 10.093 8.528 10.093 make_images_data 222 9.7 0.004 0.005 8.841 9.762 hybrid_alltoall_any 227 10.6 0.542 2.593 8.155 9.581 mp_sum_l 887 5.1 5.416 8.902 5.416 8.902 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.217 7.918 mp_irecv_dv 3311 11.0 3.197 7.865 3.197 7.865 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.122 7.158 calculate_norms 2376 9.8 6.058 6.620 6.058 6.620 multiply_cannon_sync_h2d 1332 9.7 4.856 6.387 4.856 6.387 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.904 5.150 arnoldi_extremal 4 6.8 0.000 0.000 4.664 4.692 arnoldi_normal_ev 4 7.8 0.001 0.005 4.664 4.692 build_subspace 16 8.4 0.014 0.021 4.409 4.412 ls_scf_post 1 4.0 0.003 0.023 3.899 3.919 ls_scf_store_result 1 5.0 0.000 0.000 3.438 3.539 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.139 3.367 dbcsr_matrix_vector_mult_local 304 10.0 2.745 3.222 2.747 3.224 mp_allgather_i34 111 8.7 0.839 3.080 0.839 3.080 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.592 2.665 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.121 2.491 dbcsr_data_new 4174 10.1 2.114 2.407 2.114 2.407 make_images_pack 222 9.7 1.813 2.120 1.815 2.122 dbcsr_sort_data 436 11.2 1.845 2.094 1.845 2.094 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.874 1.876 rebuild_ks_matrix 3 7.3 0.000 0.000 1.861 1.863 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.006 1.861 1.863 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.152000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1758.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_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 2.932494E+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 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.075 92.254 92.255 qs_energies 1 2.0 0.000 0.000 91.514 91.517 ls_scf 1 3.0 0.000 0.000 90.107 90.112 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.735 75.085 ls_scf_main 1 4.0 0.000 0.000 56.146 56.150 multiply_cannon 111 7.7 0.029 0.065 51.337 55.974 multiply_cannon_loop 111 8.7 0.116 0.130 48.604 52.760 density_matrix_trs4 2 5.0 0.002 0.003 50.207 50.358 mp_waitall_1 7281 11.0 22.817 33.231 22.817 33.231 ls_scf_init_scf 1 4.0 0.000 0.000 30.323 30.325 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.110 29.205 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.689 26.705 multiply_cannon_multrec 888 9.7 12.849 15.634 21.543 24.939 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.153 22.522 make_m2s 222 7.7 0.006 0.008 16.574 17.269 make_images 222 8.7 1.584 1.853 16.536 17.232 make_images_data 222 9.7 0.004 0.004 9.673 10.749 hybrid_alltoall_any 227 10.6 0.639 2.958 9.141 10.386 dbcsr_mm_accdrv_process 3754 10.4 0.296 0.505 8.228 9.522 dbcsr_mm_accdrv_process_sort 3754 11.4 7.789 9.017 7.789 9.017 mp_sum_l 887 5.1 5.139 8.536 5.139 8.536 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.420 7.298 mp_irecv_dv 2335 11.1 2.405 7.237 2.405 7.237 multiply_cannon_sync_h2d 888 9.7 6.061 7.217 6.061 7.217 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.941 6.580 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.628 6.520 arnoldi_extremal 4 6.8 0.000 0.000 5.179 5.200 arnoldi_normal_ev 4 7.8 0.001 0.005 5.178 5.200 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.869 5.100 build_subspace 16 8.4 0.014 0.020 4.870 4.876 calculate_norms 1584 9.8 4.341 4.739 4.341 4.739 mp_allgather_i34 111 8.7 0.847 3.868 0.847 3.868 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.469 3.825 ls_scf_post 1 4.0 0.000 0.000 3.639 3.643 dbcsr_matrix_vector_mult_local 304 10.0 3.027 3.616 3.029 3.618 ls_scf_store_result 1 5.0 0.000 0.000 3.375 3.462 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.852 2.947 dbcsr_data_new 4116 9.9 2.110 2.438 2.110 2.438 make_images_sizes 222 9.7 0.000 0.000 1.001 2.179 mp_alltoall_i44 222 10.7 1.001 2.178 1.001 2.178 dbcsr_sort_data 325 11.1 1.891 2.139 1.891 2.139 make_images_pack 222 9.7 1.670 1.929 1.673 1.931 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.893 1.895 rebuild_ks_matrix 3 7.3 0.000 0.000 1.875 1.878 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.875 1.877 dbcsr_finalize 304 7.8 0.026 0.032 1.600 1.863 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.255000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2299.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_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 3.309474E+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 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.063 0.268 102.573 102.574 qs_energies 1 2.0 0.008 0.067 101.522 101.666 ls_scf 1 3.0 0.003 0.021 99.791 99.876 dbcsr_multiply_generic 111 6.7 0.018 0.027 83.132 83.372 multiply_cannon 111 7.7 0.043 0.112 56.072 60.792 ls_scf_main 1 4.0 0.010 0.099 59.018 59.019 multiply_cannon_loop 111 8.7 0.153 0.170 50.954 54.595 density_matrix_trs4 2 5.0 0.003 0.014 52.831 52.959 ls_scf_init_scf 1 4.0 0.000 0.001 37.326 37.328 ls_scf_init_matrix_S 1 5.0 0.000 0.000 35.945 36.024 mp_waitall_1 6369 11.0 23.511 33.917 23.511 33.917 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 33.241 33.254 multiply_cannon_multrec 1332 9.7 14.113 17.651 22.209 25.315 make_m2s 222 7.7 0.007 0.008 21.205 22.692 make_images 222 8.7 3.148 3.603 21.155 22.644 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.911 20.255 make_images_data 222 9.7 0.004 0.004 11.830 13.502 hybrid_alltoall_any 227 10.6 0.795 3.769 11.185 12.859 dbcsr_mm_accdrv_process 3641 10.4 0.287 0.489 7.734 9.199 dbcsr_mm_accdrv_process_sort 3641 11.4 7.217 8.715 7.217 8.715 mp_sum_l 887 5.1 4.416 7.739 4.416 7.739 multiply_cannon_sync_h2d 1332 9.7 5.416 6.203 5.416 6.203 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.226 6.068 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.096 6.008 mp_irecv_dv 3229 10.9 2.070 5.921 2.070 5.921 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.596 5.709 arnoldi_extremal 4 6.8 0.000 0.000 5.282 5.297 arnoldi_normal_ev 4 7.8 0.001 0.005 5.282 5.297 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.810 4.999 build_subspace 16 8.4 0.014 0.021 4.935 4.945 compute_matrix_preconditioner 1 6.0 0.001 0.001 4.885 4.893 mp_allgather_i34 111 8.7 2.214 4.701 2.214 4.701 calculate_norms 2376 9.8 4.174 4.522 4.174 4.522 acc_transpose_blocks 1332 9.7 0.008 0.008 3.706 4.420 acc_transpose_blocks_kernels 1332 10.7 0.018 0.018 3.595 4.306 jit_kernel_transpose 1 13.0 3.577 4.289 3.577 4.289 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.625 3.889 dbcsr_matrix_vector_mult_local 304 10.0 3.197 3.689 3.199 3.691 ls_scf_post 1 4.0 0.009 0.069 3.444 3.510 dbcsr_sort_data 658 11.4 3.080 3.455 3.080 3.455 dbcsr_special_finalize 555 9.7 0.006 0.007 2.829 3.259 dbcsr_merge_single_wm 555 10.7 0.539 0.671 2.820 3.251 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.042 3.091 ls_scf_store_result 1 5.0 0.000 0.000 3.015 3.079 dbcsr_data_release 10477 10.7 1.593 2.435 1.593 2.435 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=102.574000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2712.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_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 4.741329E+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 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.050 92.322 92.323 qs_energies 1 2.0 0.000 0.000 91.507 91.512 ls_scf 1 3.0 0.000 0.000 89.571 89.575 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.948 71.144 ls_scf_main 1 4.0 0.000 0.000 56.633 56.633 multiply_cannon 111 7.7 0.084 0.144 52.586 56.084 multiply_cannon_loop 111 8.7 0.090 0.102 50.023 51.196 density_matrix_trs4 2 5.0 0.002 0.003 49.624 49.688 ls_scf_init_scf 1 4.0 0.000 0.000 29.537 29.539 mp_waitall_1 5436 11.0 24.413 28.875 24.413 28.875 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.253 28.283 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.140 26.149 multiply_cannon_multrec 444 9.7 13.775 16.369 20.947 22.957 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.520 15.252 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.324 14.789 make_m2s 222 7.7 0.005 0.005 13.673 14.613 make_images 222 8.7 2.041 2.484 13.606 14.543 hybrid_alltoall_any 227 10.6 0.801 3.825 8.182 9.859 make_images_data 222 9.7 0.003 0.004 8.381 9.653 dbcsr_mm_accdrv_process 3003 10.4 0.350 0.393 6.871 8.000 multiply_cannon_sync_h2d 444 9.7 6.707 7.896 6.707 7.896 dbcsr_mm_accdrv_process_sort 3003 11.4 6.510 7.607 6.510 7.607 arnoldi_extremal 4 6.8 0.000 0.000 5.946 5.958 arnoldi_normal_ev 4 7.8 0.002 0.005 5.945 5.958 build_subspace 16 8.4 0.015 0.019 5.558 5.567 mp_sum_l 887 5.1 2.683 4.908 2.683 4.908 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.472 4.702 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.218 4.448 dbcsr_matrix_vector_mult_local 304 10.0 3.716 4.188 3.718 4.190 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.572 4.016 mp_irecv_dv 1241 11.2 1.559 3.991 1.559 3.991 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.855 3.784 calculate_norms 792 9.8 3.607 3.744 3.607 3.744 mp_allgather_i34 111 8.7 1.150 3.662 1.150 3.662 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.518 3.571 make_images_sizes 222 9.7 0.000 0.000 0.908 3.412 mp_alltoall_i44 222 10.7 0.908 3.412 0.908 3.412 ls_scf_post 1 4.0 0.000 0.000 3.401 3.405 ls_scf_store_result 1 5.0 0.000 0.000 3.192 3.228 dbcsr_finalize 304 7.8 0.062 0.076 2.196 2.272 dbcsr_data_new 4608 9.7 1.781 2.250 1.781 2.250 dbcsr_merge_all 275 8.9 0.478 0.521 2.056 2.112 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.085 2.086 rebuild_ks_matrix 3 7.3 0.000 0.000 2.053 2.054 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.053 2.054 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 1.920 1.920 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.323000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3738.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_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 8.836542E+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 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.073 0.091 109.009 109.010 qs_energies 1 2.0 0.000 0.000 107.571 107.578 ls_scf 1 3.0 0.000 0.000 104.592 104.598 dbcsr_multiply_generic 111 6.7 0.024 0.027 78.195 78.362 ls_scf_main 1 4.0 0.000 0.000 64.245 64.246 density_matrix_trs4 2 5.0 0.002 0.003 55.206 55.265 multiply_cannon 111 7.7 0.121 0.215 51.845 53.973 multiply_cannon_loop 111 8.7 0.098 0.098 48.853 49.310 ls_scf_init_scf 1 4.0 0.000 0.000 36.526 36.527 ls_scf_init_matrix_S 1 5.0 0.000 0.000 35.040 35.055 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 32.111 32.120 mp_waitall_1 4527 11.1 22.839 26.499 22.839 26.499 make_m2s 222 7.7 0.005 0.005 22.732 23.813 make_images 222 8.7 3.585 3.886 22.624 23.704 multiply_cannon_multrec 444 9.7 17.879 18.488 22.687 23.129 hybrid_alltoall_any 227 10.6 1.657 3.640 12.858 15.844 make_images_data 222 9.7 0.003 0.004 13.100 15.263 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 11.090 13.992 multiply_cannon_sync_h2d 444 9.7 8.792 8.850 8.792 8.850 arnoldi_extremal 4 6.8 0.000 0.000 7.498 7.509 arnoldi_normal_ev 4 7.8 0.003 0.009 7.498 7.509 build_subspace 16 8.4 0.026 0.036 6.955 6.966 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.551 5.721 dbcsr_matrix_vector_mult_local 304 10.0 5.085 5.414 5.088 5.417 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.178 5.409 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.081 5.172 dbcsr_mm_accdrv_process 1814 10.4 0.276 0.591 4.624 4.916 dbcsr_mm_accdrv_process_sort 1814 11.4 4.193 4.325 4.193 4.325 ls_scf_post 1 4.0 0.000 0.000 3.821 3.828 make_images_sizes 222 9.7 0.000 0.000 1.481 3.678 mp_alltoall_i44 222 10.7 1.480 3.678 1.480 3.678 mp_allgather_i34 111 8.7 1.104 3.623 1.104 3.623 ls_scf_store_result 1 5.0 0.000 0.000 3.538 3.575 compute_matrix_preconditioner 1 6.0 0.002 0.002 3.296 3.302 calculate_norms 792 9.8 3.231 3.270 3.231 3.270 dbcsr_finalize 304 7.8 0.082 0.089 3.084 3.175 dbcsr_merge_all 275 8.9 0.887 0.921 2.867 2.953 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.946 2.947 dbcsr_complete_redistribute 5 7.6 1.437 1.485 2.757 2.865 acc_transpose_blocks 444 9.7 0.003 0.003 2.507 2.838 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 2.425 2.756 jit_kernel_transpose 1 13.0 2.419 2.750 2.419 2.750 matrix_ls_to_qs 2 6.0 0.000 0.000 2.412 2.526 dbcsr_sort_data 325 11.1 2.446 2.503 2.446 2.503 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.381 2.383 dbcsr_data_new 6591 9.6 1.914 2.378 1.914 2.378 dbcsr_new_transposed 4 7.5 0.242 0.253 2.316 2.328 rebuild_ks_matrix 3 7.3 0.000 0.000 2.315 2.317 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.315 2.317 dbcsr_add_d 103 6.2 0.000 0.000 2.142 2.224 dbcsr_add_anytype 103 7.2 0.858 0.890 2.142 2.224 dbcsr_frobenius_norm 74 6.6 2.055 2.126 2.180 2.222 dbcsr_data_release 12724 10.6 1.978 2.201 1.978 2.201 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.010000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7085.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1eb99f0345e8f3b2962f17f4aadd5584dc421ad2_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 588.984320E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 4805170. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.112 241.414 241.418 qs_mol_dyn_low 1 2.0 0.003 0.017 239.991 240.012 qs_forces 5 3.8 0.004 0.005 239.900 239.903 qs_energies 5 4.8 0.001 0.002 236.328 236.353 scf_env_do_scf 5 5.8 0.000 0.001 219.933 219.936 scf_env_do_scf_inner_loop 105 6.6 0.002 0.008 192.394 192.396 qs_scf_new_mos 105 7.6 0.000 0.001 150.106 150.338 qs_scf_loop_do_ot 105 8.6 0.001 0.001 150.106 150.338 ot_scf_mini 105 9.6 0.003 0.003 140.199 140.420 dbcsr_multiply_generic 1445 12.2 0.127 0.134 136.266 136.789 multiply_cannon 1445 13.2 0.275 0.286 116.657 119.335 multiply_cannon_loop 1445 14.2 2.826 2.968 114.790 116.403 velocity_verlet 4 3.0 0.004 0.019 108.833 108.834 ot_mini 105 10.6 0.001 0.001 61.037 61.164 qs_ot_get_p 112 10.4 0.001 0.001 49.018 49.334 multiply_cannon_multrec 69360 15.2 29.629 34.408 39.710 44.557 mp_waitall_1 488190 16.1 37.017 43.575 37.017 43.575 qs_ot_get_derivative 55 11.6 0.001 0.001 39.187 39.348 qs_ot_p2m_diag 40 11.0 0.020 0.030 37.525 37.579 cp_dbcsr_syevd 40 12.0 0.002 0.002 34.262 34.264 multiply_cannon_metrocomm3 69360 15.2 0.200 0.215 26.695 33.998 multiply_cannon_sync_h2d 69360 15.2 29.053 33.248 29.053 33.248 rebuild_ks_matrix 110 8.4 0.000 0.000 30.129 30.306 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.026 30.129 30.306 cp_fm_syevd 40 13.0 0.000 0.000 28.826 28.959 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.686 27.837 init_scf_loop 7 6.6 0.000 0.000 27.491 27.492 cp_fm_redistribute_end 40 14.0 11.932 23.801 11.937 23.803 cp_fm_syevd_base 40 14.0 11.856 23.727 11.856 23.727 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.112 23.383 apply_single 62 13.6 0.000 0.000 23.112 23.382 prepare_preconditioner 7 7.6 0.000 0.000 22.502 22.531 make_preconditioner 7 8.6 0.000 0.000 22.502 22.531 ot_new_cg_direction 55 11.6 0.001 0.001 21.125 21.126 qs_rho_update_rho_low 110 7.6 0.001 0.001 18.135 18.459 calculate_rho_elec 110 8.6 0.029 0.032 18.135 18.458 make_full_inverse_cholesky 7 9.6 0.000 0.000 15.015 15.080 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.875 15.047 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.040 14.188 init_scf_run 5 5.8 0.000 0.000 13.770 13.771 scf_env_initial_rho_setup 5 6.8 0.001 0.001 13.770 13.770 mp_sum_l 4764 12.2 12.355 13.323 12.355 13.323 pw_transfer 1645 12.4 0.079 0.103 12.607 12.897 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 12.468 12.763 density_rs2pw 110 9.6 0.005 0.006 12.194 12.705 calculate_dm_sparse 110 9.5 0.000 0.001 11.582 11.831 fft_wrap_pw1pw2_240 915 15.0 1.129 1.222 10.990 11.237 dbcsr_mm_accdrv_process 154766 15.8 6.192 6.407 9.950 11.201 qs_vxc_create 110 10.4 0.002 0.003 10.949 10.999 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.724 10.806 cp_fm_cholesky_invert 7 10.6 10.386 10.397 10.386 10.397 check_diag 80 13.5 8.683 8.936 9.957 10.091 fft3d_pb 915 16.0 2.381 2.660 9.224 9.495 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 9.191 9.207 sum_up_and_integrate 60 10.3 0.001 0.002 9.080 9.102 integrate_v_rspace 60 11.3 0.001 0.002 9.063 9.085 transfer_rs2pw 445 10.6 0.007 0.008 8.567 9.070 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 8.732 8.796 acc_transpose_blocks 69360 15.2 0.350 0.370 8.179 8.787 calculate_first_density_matrix 1 7.0 0.000 0.008 8.739 8.760 multiply_cannon_metrocomm1 69360 15.2 0.094 0.102 5.211 8.676 xc_rho_set_and_dset_create 110 12.4 0.076 0.097 7.645 7.910 make_m2s 2890 13.2 0.079 0.088 6.851 7.483 make_images 2890 14.2 0.240 0.261 6.743 7.376 xc_vxc_pw_create 60 11.3 0.039 0.050 7.308 7.357 make_full_single_inverse 7 9.6 0.001 0.001 7.148 7.178 xc_pw_derive 510 13.4 0.005 0.006 6.662 6.752 mp_alltoall_z22v 2340 17.7 6.307 6.567 6.307 6.567 acc_transpose_blocks_kernels 69360 16.2 0.852 0.897 5.442 5.890 potential_pw2rs 60 12.3 0.003 0.003 5.239 5.267 mp_waitany 7680 13.5 4.417 5.123 4.417 5.123 jit_kernel_transpose 5 15.0 4.590 5.015 4.590 5.015 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=241.418000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=560.800000, yerr=1.600000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 1eb99f0345e8f3b2962f17f4aadd5584dc421ad2 Summary: empty Status: OK