OPENMPI detected AcceleratorCudaInit[0]: ======================== AcceleratorCudaInit[0]: Device Number : 0 AcceleratorCudaInit[0]: ======================== AcceleratorCudaInit[0]: Device identifier: NVIDIA A100-SXM4-40GB AcceleratorCudaInit[0]: totalGlobalMem: 42505273344 AcceleratorCudaInit[0]: managedMemory: 1 AcceleratorCudaInit[0]: isMultiGpuBoard: 0 AcceleratorCudaInit[0]: warpSize: 32 AcceleratorCudaInit[0]: pciBusID: 3 AcceleratorCudaInit[0]: pciDeviceID: 0 AcceleratorCudaInit[0]: maxGridSize (2147483647,65535,65535) AcceleratorCudaInit: using default device AcceleratorCudaInit: assume user either uses a) IBM jsrun, or AcceleratorCudaInit: b) invokes through a wrapping script to set CUDA_VISIBLE_DEVICES, UCX_NET_DEVICES, and numa binding AcceleratorCudaInit: Configure options --enable-summit, --enable-select-gpu=no AcceleratorCudaInit: ================================================ SharedMemoryMpi: World communicator of size 16 SharedMemoryMpi: Node communicator of size 4 0SharedMemoryMpi: SharedMemoryMPI.cc acceleratorAllocDevice 2147483648bytes at 0x14e9c0000000 for comms buffers Setting up IPC __|__|__|__|__|__|__|__|__|__|__|__|__|__|__ __|__|__|__|__|__|__|__|__|__|__|__|__|__|__ __|_ | | | | | | | | | | | | _|__ __|_ _|__ __|_ GGGG RRRR III DDDD _|__ __|_ G R R I D D _|__ __|_ G R R I D D _|__ __|_ G GG RRRR I D D _|__ __|_ G G R R I D D _|__ __|_ GGGG R R III DDDD _|__ __|_ _|__ __|__|__|__|__|__|__|__|__|__|__|__|__|__|__ __|__|__|__|__|__|__|__|__|__|__|__|__|__|__ | | | | | | | | | | | | | | Copyright (C) 2015 Peter Boyle, Azusa Yamaguchi, Guido Cossu, Antonin Portelli and other authors This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. Current Grid git commit hash=e188c0512ebee79bfb15906676af1c9e142aa21a: (HEAD -> develop) uncommited changes Grid : Message : ================================================ Grid : Message : MPI is initialised and logging filters activated Grid : Message : ================================================ Grid : Message : Requested 2147483648 byte stencil comms buffers Grid : Message : MemoryManager Cache 34004218675 bytes Grid : Message : MemoryManager::Init() setting up Grid : Message : MemoryManager::Init() cache pool for recent allocations: SMALL 32 LARGE 8 Grid : Message : MemoryManager::Init() Non unified: Caching accelerator data in dedicated memory Grid : Message : MemoryManager::Init() Using cudaMalloc Grid : Message : 0.717713 s : Grid Layout Grid : Message : 0.717716 s : Global lattice size : 64 64 64 64 Grid : Message : 0.717724 s : OpenMP threads : 4 Grid : Message : 0.717725 s : MPI tasks : 2 2 2 2 Grid : Message : 0.801634 s : Making s innermost grids Grid : Message : 0.844903 s : Initialising 4d RNG Grid : Message : 0.940001 s : Intialising parallel RNG with unique string 'The 4D RNG' Grid : Message : 0.940060 s : Seed SHA256: 49db4542db694e3b1a74bf2592a8c1b83bfebbe18401693c2609a4c3af1 Grid : Message : 1.338368 s : Initialising 5d RNG Grid : Message : 2.859273 s : Intialising parallel RNG with unique string 'The 5D RNG' Grid : Message : 2.859304 s : Seed SHA256: b6316f2fac44ce14111f93e0296389330b077bfd0a7b359f781c58589f8a Grid : Message : 11.140924 s : Initialised RNGs Grid : Message : 13.433456 s : Drawing gauge field Grid : Message : 13.955847 s : Random gauge initialised Grid : Message : 15.528535 s : Setting up Cshift based reference Grid : Message : 21.484340 s : ***************************************************************** Grid : Message : 21.484840 s : * Kernel options --dslash-generic, --dslash-unroll, --dslash-asm Grid : Message : 21.484860 s : ***************************************************************** Grid : Message : 21.484870 s : ***************************************************************** Grid : Message : 21.484880 s : * Benchmarking DomainWallFermionR::Dhop Grid : Message : 21.484890 s : * Vectorising space-time by 8 Grid : Message : 21.484900 s : * VComplexF size is 64 B Grid : Message : 21.484910 s : * SINGLE precision Grid : Message : 21.484920 s : * Using Overlapped Comms/Compute Grid : Message : 21.484930 s : * Using GENERIC Nc WilsonKernels Grid : Message : 21.484940 s : ***************************************************************** Grid : Message : 22.344741 s : Called warmup Grid : Message : 49.832292 s : Called Dw 3000 times in 2.74873e+07 us Grid : Message : 49.832358 s : mflop/s = 3.86726e+07 Grid : Message : 49.832360 s : mflop/s per rank = 2.41704e+06 Grid : Message : 49.832361 s : mflop/s per node = 9.66814e+06 Grid : Message : 49.832362 s : RF GiB/s (base 2) = 78581.7 Grid : Message : 49.832363 s : mem GiB/s (base 2) = 49113.6 Grid : Message : 49.835924 s : norm diff 1.03481e-13 Grid : Message : 49.870568 s : #### Dhop calls report Grid : Message : 49.870574 s : WilsonFermion5D Number of DhopEO Calls : 6002 Grid : Message : 49.870598 s : WilsonFermion5D TotalTime /Calls : 4616.79 us Grid : Message : 49.870600 s : WilsonFermion5D CommTime /Calls : 3241.77 us Grid : Message : 49.870601 s : WilsonFermion5D FaceTime /Calls : 469.006 us Grid : Message : 49.870602 s : WilsonFermion5D ComputeTime1/Calls : 27.0492 us Grid : Message : 49.870603 s : WilsonFermion5D ComputeTime2/Calls : 926.33 us Grid : Message : 49.870614 s : Average mflops/s per call : 6.71631e+09 Grid : Message : 49.870619 s : Average mflops/s per call per rank : 4.19769e+08 Grid : Message : 49.870621 s : Average mflops/s per call per node : 1.67908e+09 Grid : Message : 49.870626 s : Average mflops/s per call (full) : 3.90723e+07 Grid : Message : 49.870627 s : Average mflops/s per call per rank (full): 2.44202e+06 Grid : Message : 49.870628 s : Average mflops/s per call per node (full): 9.76808e+06 Grid : Message : 49.870629 s : WilsonFermion5D Stencil Grid : Message : 49.870630 s : WilsonFermion5D StencilEven Grid : Message : 49.870631 s : WilsonFermion5D StencilOdd Grid : Message : 49.870632 s : WilsonFermion5D Stencil Reporti() Grid : Message : 49.870633 s : WilsonFermion5D StencilEven Reporti() Grid : Message : 49.870634 s : WilsonFermion5D StencilOdd Reporti() Grid : Message : 77.321890 s : Compare to naive wilson implementation Dag to verify correctness Grid : Message : 77.321911 s : Called DwDag Grid : Message : 77.321912 s : norm dag result 12.0421 Grid : Message : 77.334619 s : norm dag ref 12.0421 Grid : Message : 77.350515 s : norm dag diff 7.63236e-14 Grid : Message : 77.389923 s : Calling Deo and Doe and //assert Deo+Doe == Dunprec Grid : Message : 77.769815 s : src_e0.499997 Grid : Message : 77.847560 s : src_o0.500003 Grid : Message : 77.917493 s : ********************************************************* Grid : Message : 77.917496 s : * Benchmarking DomainWallFermionF::DhopEO Grid : Message : 77.917497 s : * Vectorising space-time by 8 Grid : Message : 77.917498 s : * SINGLE precision Grid : Message : 77.917499 s : * Using Overlapped Comms/Compute Grid : Message : 77.917500 s : * Using GENERIC Nc WilsonKernels Grid : Message : 77.917501 s : ********************************************************* Grid : Message : 91.412946 s : Deo mflop/s = 3.95925e+07 Grid : Message : 91.412978 s : Deo mflop/s per rank 2.47453e+06 Grid : Message : 91.412980 s : Deo mflop/s per node 9.89813e+06 Grid : Message : 91.412983 s : #### Dhop calls report Grid : Message : 91.412984 s : WilsonFermion5D Number of DhopEO Calls : 3001 Grid : Message : 91.412986 s : WilsonFermion5D TotalTime /Calls : 4496.84 us Grid : Message : 91.412988 s : WilsonFermion5D CommTime /Calls : 3057.28 us Grid : Message : 91.412989 s : WilsonFermion5D FaceTime /Calls : 528.499 us Grid : Message : 91.412990 s : WilsonFermion5D ComputeTime1/Calls : 16.1939 us Grid : Message : 91.412991 s : WilsonFermion5D ComputeTime2/Calls : 942.557 us Grid : Message : 91.413021 s : Average mflops/s per call : 1.12574e+10 Grid : Message : 91.413023 s : Average mflops/s per call per rank : 7.03586e+08 Grid : Message : 91.413024 s : Average mflops/s per call per node : 2.81434e+09 Grid : Message : 91.413025 s : Average mflops/s per call (full) : 4.01145e+07 Grid : Message : 91.413026 s : Average mflops/s per call per rank (full): 2.50716e+06 Grid : Message : 91.413027 s : Average mflops/s per call per node (full): 1.00286e+07 Grid : Message : 91.413028 s : WilsonFermion5D Stencil Grid : Message : 91.413029 s : WilsonFermion5D StencilEven Grid : Message : 91.413030 s : WilsonFermion5D StencilOdd Grid : Message : 91.413031 s : WilsonFermion5D Stencil Reporti() Grid : Message : 91.413032 s : WilsonFermion5D StencilEven Reporti() Grid : Message : 91.413033 s : WilsonFermion5D StencilOdd Reporti() Grid : Message : 91.470394 s : r_e6.02111 Grid : Message : 91.476539 s : r_o6.02102 Grid : Message : 91.482442 s : res12.0421 Grid : Message : 92.138799 s : norm diff 0 Grid : Message : 92.545354 s : norm diff even 0 Grid : Message : 92.619444 s : norm diff odd 0