tu-c0r1n72 - 1 device=1 binding=--interleave=2,3 tu-c0r1n72 - 0 device=0 binding=--interleave=0,1 tu-c0r1n72 - 3 device=3 binding=--interleave=6,7 tu-c0r1n72 - 2 device=2 binding=--interleave=4,5 tu-c0r1n84 - 2 device=2 binding=--interleave=4,5 tu-c0r1n84 - 3 device=3 binding=--interleave=6,7 tu-c0r1n84 - 1 device=1 binding=--interleave=2,3 tu-c0r1n84 - 0 device=0 binding=--interleave=0,1 tu-c0r1n75 - 1 device=1 binding=--interleave=2,3 tu-c0r1n93 - 0 device=0 binding=--interleave=0,1 tu-c0r1n75 - 0 device=0 binding=--interleave=0,1 tu-c0r1n93 - 2 device=2 binding=--interleave=4,5 tu-c0r1n75 - 2 device=2 binding=--interleave=4,5 tu-c0r1n75 - 3 device=3 binding=--interleave=6,7 tu-c0r1n93 - 1 device=1 binding=--interleave=2,3 tu-c0r1n93 - 3 device=3 binding=--interleave=6,7 tu-c0r1n87 - 1 device=1 binding=--interleave=2,3 tu-c0r1n87 - 3 device=3 binding=--interleave=6,7 tu-c0r1n87 - 0 device=0 binding=--interleave=0,1 tu-c0r1n87 - 2 device=2 binding=--interleave=4,5 tu-c0r1n81 - 0 device=0 binding=--interleave=0,1 tu-c0r1n81 - 1 device=1 binding=--interleave=2,3 tu-c0r1n81 - 3 device=3 binding=--interleave=6,7 tu-c0r1n81 - 2 device=2 binding=--interleave=4,5 tu-c0r1n90 - 2 device=2 binding=--interleave=4,5 tu-c0r1n90 - 1 device=1 binding=--interleave=2,3 tu-c0r1n90 - 3 device=3 binding=--interleave=6,7 tu-c0r1n90 - 0 device=0 binding=--interleave=0,1 tu-c0r1n78 - 0 device=0 binding=--interleave=0,1 tu-c0r1n78 - 1 device=1 binding=--interleave=2,3 tu-c0r1n78 - 2 device=2 binding=--interleave=4,5 tu-c0r1n78 - 3 device=3 binding=--interleave=6,7 OPENMPI detected AcceleratorCudaInit: using default device AcceleratorCudaInit: assume user either uses AcceleratorCudaInit: 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 OPENMPI detected AcceleratorCudaInit: using default device AcceleratorCudaInit: assume user either uses AcceleratorCudaInit: 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 OPENMPI detected AcceleratorCudaInit: using default device AcceleratorCudaInit: assume user either uses AcceleratorCudaInit: 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 OPENMPI detected AcceleratorCudaInit: using default device AcceleratorCudaInit: assume user either uses AcceleratorCudaInit: 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 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 AcceleratorCudaInit: 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 OPENMPI detected AcceleratorCudaInit: using default device AcceleratorCudaInit: assume user either uses AcceleratorCudaInit: 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 OPENMPI detected AcceleratorCudaInit: using default device AcceleratorCudaInit: assume user either uses AcceleratorCudaInit: 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 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 AcceleratorCudaInit: 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 local rank 0 device 0 bus id: 0000:03:00.0 AcceleratorCudaInit: ================================================ AcceleratorCudaInit: ================================================ AcceleratorCudaInit: ================================================ AcceleratorCudaInit: ================================================ local rank 2 device 0 bus id: 0000:84:00.0 AcceleratorCudaInit: ================================================ local rank 0 device 0 bus id: 0000:03:00.0 AcceleratorCudaInit: ================================================ AcceleratorCudaInit: ================================================ AcceleratorCudaInit: ================================================ local rank 1 device 0 bus id: 0000:44:00.0 local rank 3 device 0 bus id: 0000:C4:00.0 SharedMemoryMpi: World communicator of size 32 SharedMemoryMpi: Node communicator of size 4 0SharedMemoryMpi: SharedMemoryMPI.cc acceleratorAllocDevice 2147483648bytes at 0x154140000000 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=188d2c7a4dc77807b545f5f2813cdb589b9e44ca: (HEAD -> develop, gh/develop, gh/HEAD) 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 8 LARGE 2 Grid : Message : MemoryManager::Init() Non unified: Caching accelerator data in dedicated memory Grid : Message : MemoryManager::Init() Using cudaMalloc Grid : Message : 1.483229 s : Grid Layout Grid : Message : 1.483233 s : Global lattice size : 64 64 64 128 Grid : Message : 1.483238 s : OpenMP threads : 4 Grid : Message : 1.483241 s : MPI tasks : 2 2 2 4 Grid : Message : 1.521315 s : Making s innermost grids Grid : Message : 1.579588 s : Initialising 4d RNG Grid : Message : 1.674602 s : Intialising parallel RNG with unique string 'The 4D RNG' Grid : Message : 1.674630 s : Seed SHA256: 49db4542db694e3b1a74bf2592a8c1b83bfebbe18401693c2609a4c3af1 Grid : Message : 2.721485 s : Initialising 5d RNG Grid : Message : 4.155442 s : Intialising parallel RNG with unique string 'The 5D RNG' Grid : Message : 4.155467 s : Seed SHA256: b6316f2fac44ce14111f93e0296389330b077bfd0a7b359f781c58589f8a Grid : Message : 20.834199 s : Initialised RNGs Grid : Message : 25.213264 s : Drawing gauge field Grid : Message : 25.744675 s : Random gauge initialised Grid : Message : 25.758258 s : Setting up Cshift based reference Grid : Message : 55.878956 s : ***************************************************************** Grid : Message : 55.878977 s : * Kernel options --dslash-generic, --dslash-unroll, --dslash-asm Grid : Message : 55.878978 s : ***************************************************************** Grid : Message : 55.878979 s : ***************************************************************** Grid : Message : 55.878980 s : * Benchmarking DomainWallFermionR::Dhop Grid : Message : 55.878981 s : * Vectorising space-time by 8 Grid : Message : 55.878982 s : * VComplexF size is 64 B Grid : Message : 55.878983 s : * SINGLE precision Grid : Message : 55.878985 s : * Using Overlapped Comms/Compute Grid : Message : 55.878986 s : * Using GENERIC Nc WilsonKernels Grid : Message : 55.878987 s : ***************************************************************** Grid : Message : 57.921484 s : Called warmup Grid : Message : 442.780552 s : Called Dw 30000 times in 3.84858e+08 us Grid : Message : 442.780604 s : mflop/s = 5.52414e+07 Grid : Message : 442.780606 s : mflop/s per rank = 1.72629e+06 Grid : Message : 442.780608 s : mflop/s per node = 6.90517e+06 Grid : Message : 442.780610 s : RF GiB/s (base 2) = 112249 Grid : Message : 442.780612 s : mem GiB/s (base 2) = 70155.8 Grid : Message : 442.784549 s : norm diff 1.07359e-13 Grid : Message : 442.846713 s : #### Dhop calls report Grid : Message : 442.846719 s : WilsonFermion5D Number of DhopEO Calls : 60002 Grid : Message : 442.846722 s : WilsonFermion5D TotalTime /Calls : 6418.16 us Grid : Message : 442.846724 s : WilsonFermion5D CommTime /Calls : 4760.19 us Grid : Message : 442.846726 s : WilsonFermion5D FaceTime /Calls : 658.377 us Grid : Message : 442.846728 s : WilsonFermion5D ComputeTime1/Calls : 4.66556 us Grid : Message : 442.846730 s : WilsonFermion5D ComputeTime2/Calls : 1014.28 us Grid : Message : 442.846825 s : Average mflops/s per call : 6.36075e+10 Grid : Message : 442.846829 s : Average mflops/s per call per rank : 1.98773e+09 Grid : Message : 442.846831 s : Average mflops/s per call per node : 7.95094e+09 Grid : Message : 442.846833 s : Average mflops/s per call (full) : 5.6212e+07 Grid : Message : 442.846835 s : Average mflops/s per call per rank (full): 1.75662e+06 Grid : Message : 442.846840 s : Average mflops/s per call per node (full): 7.0265e+06 Grid : Message : 442.846843 s : WilsonFermion5D Stencil Grid : Message : 442.846844 s : WilsonFermion5D StencilEven Grid : Message : 442.846845 s : WilsonFermion5D StencilOdd Grid : Message : 442.846847 s : WilsonFermion5D Stencil Reporti() Grid : Message : 442.846850 s : WilsonFermion5D StencilEven Reporti() Grid : Message : 442.846852 s : WilsonFermion5D StencilOdd Reporti() Grid : Message : 499.244310 s : Compare to naive wilson implementation Dag to verify correctness Grid : Message : 499.244690 s : Called DwDag Grid : Message : 499.244700 s : norm dag result 12.0421 Grid : Message : 499.504590 s : norm dag ref 12.0421 Grid : Message : 499.714130 s : norm dag diff 7.28475e-14 Grid : Message : 499.123008 s : Calling Deo and Doe and //assert Deo+Doe == Dunprec Grid : Message : 499.503539 s : src_e0.499997 Grid : Message : 499.909722 s : src_o0.500003 Grid : Message : 500.190490 s : ********************************************************* Grid : Message : 500.190540 s : * Benchmarking DomainWallFermionF::DhopEO Grid : Message : 500.190550 s : * Vectorising space-time by 8 Grid : Message : 500.190560 s : * SINGLE precision Grid : Message : 500.190570 s : * Using Overlapped Comms/Compute Grid : Message : 500.190580 s : * Using GENERIC Nc WilsonKernels Grid : Message : 500.190590 s : ********************************************************* Grid : Message : 699.497514 s : Deo mflop/s = 5.33023e+07 Grid : Message : 699.497545 s : Deo mflop/s per rank 1.6657e+06 Grid : Message : 699.497548 s : Deo mflop/s per node 6.66279e+06 Grid : Message : 699.497554 s : #### Dhop calls report Grid : Message : 699.497558 s : WilsonFermion5D Number of DhopEO Calls : 30001 Grid : Message : 699.497561 s : WilsonFermion5D TotalTime /Calls : 6648.94 us Grid : Message : 699.497565 s : WilsonFermion5D CommTime /Calls : 4849.76 us Grid : Message : 699.497568 s : WilsonFermion5D FaceTime /Calls : 784.005 us Grid : Message : 699.497572 s : WilsonFermion5D ComputeTime1/Calls : 6.0565 us Grid : Message : 699.497574 s : WilsonFermion5D ComputeTime2/Calls : 1042.85 us Grid : Message : 699.497596 s : Average mflops/s per call : 5.10455e+10 Grid : Message : 699.497600 s : Average mflops/s per call per rank : 1.59517e+09 Grid : Message : 699.497603 s : Average mflops/s per call per node : 6.38068e+09 Grid : Message : 699.497607 s : Average mflops/s per call (full) : 5.42608e+07 Grid : Message : 699.497609 s : Average mflops/s per call per rank (full): 1.69565e+06 Grid : Message : 699.497612 s : Average mflops/s per call per node (full): 6.78261e+06 Grid : Message : 699.497615 s : WilsonFermion5D Stencil Grid : Message : 699.497617 s : WilsonFermion5D StencilEven Grid : Message : 699.497620 s : WilsonFermion5D StencilOdd Grid : Message : 699.497623 s : WilsonFermion5D Stencil Reporti() Grid : Message : 699.497625 s : WilsonFermion5D StencilEven Reporti() Grid : Message : 699.497627 s : WilsonFermion5D StencilOdd Reporti() Grid : Message : 699.575273 s : r_e6.02113 Grid : Message : 699.587804 s : r_o6.02101 Grid : Message : 699.596579 s : res12.0421 Grid : Message : 700.295444 s : norm diff 0 Grid : Message : 701.307610 s : norm diff even 0 Grid : Message : 701.415907 s : norm diff odd 0