SLURM detected AcceleratorCudaInit[0]: ======================== AcceleratorCudaInit[0]: Device Number : 0 AcceleratorCudaInit[0]: ======================== AcceleratorCudaInit[0]: Device identifier: A100-SXM4-40GB AcceleratorCudaInit[0]: totalGlobalMem: 42506321920 AcceleratorCudaInit[0]: managedMemory: 1 AcceleratorCudaInit[0]: isMultiGpuBoard: 0 AcceleratorCudaInit[0]: warpSize: 32 AcceleratorCudaInit[0]: pciBusID: 2 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-setdevice=no AcceleratorCudaInit: ================================================ SharedMemoryMpi: World communicator of size 16 SharedMemoryMpi: Node communicator of size 4 0SharedMemoryMpi: SharedMemoryMPI.cc acceleratorAllocDevice 2147483648bytes at 0x7fbae0000000 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=b2ccaad761798e93a9314f97d8a4d1f851c6962a: (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 34005057536 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 : 0.692368 s : Grid Layout Grid : Message : 0.692369 s : Global lattice size : 48 48 48 48 Grid : Message : 0.692372 s : OpenMP threads : 32 Grid : Message : 0.692372 s : MPI tasks : 2 2 2 2 Grid : Message : 0.701977 s : Making s innermost grids Grid : Message : 0.711295 s : Initialising 4d RNG Grid : Message : 0.734938 s : Intialising parallel RNG with unique string 'The 4D RNG' Grid : Message : 0.734948 s : Seed SHA256: 49db4542db694e3b1a74bf2592a8c1b83bfebbe18401693c2609a4c3af1 Grid : Message : 0.798281 s : Initialising 5d RNG Grid : Message : 1.161711 s : Intialising parallel RNG with unique string 'The 5D RNG' Grid : Message : 1.161728 s : Seed SHA256: b6316f2fac44ce14111f93e0296389330b077bfd0a7b359f781c58589f8a Grid : Message : 1.522440 s : Initialised RNGs Grid : Message : 2.260710 s : Drawing gauge field Grid : Message : 2.102597 s : Random gauge initialised Grid : Message : 2.562592 s : Setting up Cshift based reference Grid : Message : 3.121880 s : ***************************************************************** Grid : Message : 3.121970 s : * Kernel options --dslash-generic, --dslash-unroll, --dslash-asm Grid : Message : 3.121980 s : ***************************************************************** Grid : Message : 3.121980 s : ***************************************************************** Grid : Message : 3.121980 s : * Benchmarking DomainWallFermionR::Dhop Grid : Message : 3.121980 s : * Vectorising space-time by 8 Grid : Message : 3.121980 s : * VComplexF size is 64 B Grid : Message : 3.121990 s : * SINGLE precision Grid : Message : 3.121990 s : * Using Overlapped Comms/Compute Grid : Message : 3.121990 s : * Using GENERIC Nc WilsonKernels Grid : Message : 3.121990 s : ***************************************************************** Grid : Message : 3.350688 s : Called warmup Grid : Message : 35.847527 s : Called Dw 3000 times in 3.24968e+07 us Grid : Message : 35.847576 s : mflop/s = 1.035e+07 Grid : Message : 35.847578 s : mflop/s per rank = 646874 Grid : Message : 35.847579 s : mflop/s per node = 2.5875e+06 Grid : Message : 35.847580 s : RF GiB/s (base 2) = 21030.9 Grid : Message : 35.847581 s : mem GiB/s (base 2) = 13144.3 Grid : Message : 35.848697 s : norm diff 1.03662e-13 Grid : Message : 35.861967 s : #### Dhop calls report Grid : Message : 35.861973 s : WilsonFermion5D Number of DhopEO Calls : 6002 Grid : Message : 35.861976 s : WilsonFermion5D TotalTime /Calls : 5426 us Grid : Message : 35.861977 s : WilsonFermion5D CommTime /Calls : 4817.47 us Grid : Message : 35.861978 s : WilsonFermion5D FaceTime /Calls : 246.175 us Grid : Message : 35.861979 s : WilsonFermion5D ComputeTime1/Calls : 8.72676 us Grid : Message : 35.861980 s : WilsonFermion5D ComputeTime2/Calls : 370.494 us Grid : Message : 35.861995 s : Average mflops/s per call : 6.50606e+09 Grid : Message : 35.861999 s : Average mflops/s per call per rank : 4.06629e+08 Grid : Message : 35.862000 s : Average mflops/s per call per node : 1.62652e+09 Grid : Message : 35.862001 s : Average mflops/s per call (full) : 1.0519e+07 Grid : Message : 35.862002 s : Average mflops/s per call per rank (full): 657438 Grid : Message : 35.862003 s : Average mflops/s per call per node (full): 2.62975e+06 Grid : Message : 35.862004 s : WilsonFermion5D Stencil Grid : Message : 35.862004 s : WilsonFermion5D StencilEven Grid : Message : 35.862004 s : WilsonFermion5D StencilOdd Grid : Message : 35.862004 s : WilsonFermion5D Stencil Reporti() Grid : Message : 35.862004 s : WilsonFermion5D StencilEven Reporti() Grid : Message : 35.862004 s : WilsonFermion5D StencilOdd Reporti() Grid : Message : 39.599406 s : Compare to naive wilson implementation Dag to verify correctness Grid : Message : 39.599421 s : Called DwDag Grid : Message : 39.599422 s : norm dag result 12.0422 Grid : Message : 39.604317 s : norm dag ref 12.0422 Grid : Message : 39.609961 s : norm dag diff 7.6086e-14 Grid : Message : 39.624145 s : Calling Deo and Doe and //assert Deo+Doe == Dunprec Grid : Message : 39.772334 s : src_e0.499992 Grid : Message : 39.795705 s : src_o0.500008 Grid : Message : 39.816822 s : ********************************************************* Grid : Message : 39.816824 s : * Benchmarking DomainWallFermionF::DhopEO Grid : Message : 39.816825 s : * Vectorising space-time by 8 Grid : Message : 39.816826 s : * SINGLE precision Grid : Message : 39.816827 s : * Using Overlapped Comms/Compute Grid : Message : 39.816828 s : * Using GENERIC Nc WilsonKernels Grid : Message : 39.816828 s : ********************************************************* Grid : Message : 56.382758 s : Deo mflop/s = 1.017e+07 Grid : Message : 56.382779 s : Deo mflop/s per rank 635627 Grid : Message : 56.382781 s : Deo mflop/s per node 2.54251e+06 Grid : Message : 56.382783 s : #### Dhop calls report Grid : Message : 56.382784 s : WilsonFermion5D Number of DhopEO Calls : 3001 Grid : Message : 56.382785 s : WilsonFermion5D TotalTime /Calls : 5519.98 us Grid : Message : 56.382786 s : WilsonFermion5D CommTime /Calls : 4856.39 us Grid : Message : 56.382787 s : WilsonFermion5D FaceTime /Calls : 303.043 us Grid : Message : 56.382788 s : WilsonFermion5D ComputeTime1/Calls : 6.77807 us Grid : Message : 56.382789 s : WilsonFermion5D ComputeTime2/Calls : 376.551 us Grid : Message : 56.382810 s : Average mflops/s per call : 8.31124e+09 Grid : Message : 56.382811 s : Average mflops/s per call per rank : 5.19453e+08 Grid : Message : 56.382812 s : Average mflops/s per call per node : 2.07781e+09 Grid : Message : 56.382813 s : Average mflops/s per call (full) : 1.03399e+07 Grid : Message : 56.382814 s : Average mflops/s per call per rank (full): 646244 Grid : Message : 56.382815 s : Average mflops/s per call per node (full): 2.58498e+06 Grid : Message : 56.382816 s : WilsonFermion5D Stencil Grid : Message : 56.382816 s : WilsonFermion5D StencilEven Grid : Message : 56.382816 s : WilsonFermion5D StencilOdd Grid : Message : 56.382816 s : WilsonFermion5D Stencil Reporti() Grid : Message : 56.382816 s : WilsonFermion5D StencilEven Reporti() Grid : Message : 56.382816 s : WilsonFermion5D StencilOdd Reporti() Grid : Message : 56.414571 s : r_e6.02129 Grid : Message : 56.417837 s : r_o6.02097 Grid : Message : 56.420535 s : res12.0423 Grid : Message : 56.611957 s : norm diff 0 Grid : Message : 56.730597 s : norm diff even 0 Grid : Message : 56.752566 s : norm diff odd 0