tu-c0r2n72 - 0 device=0 binding=--interleave=0,1 tu-c0r2n72 - 1 device=1 binding=--interleave=2,3 tu-c0r2n72 - 2 device=2 binding=--interleave=4,5 tu-c0r2n72 - 3 device=3 binding=--interleave=6,7 tu-c0r2n75 - 1 device=1 binding=--interleave=2,3 tu-c0r2n75 - 3 device=3 binding=--interleave=6,7 tu-c0r2n75 - 2 device=2 binding=--interleave=4,5 tu-c0r2n78 - 0 device=0 binding=--interleave=0,1 tu-c0r2n75 - 0 device=0 binding=--interleave=0,1 tu-c0r2n78 - 1 device=1 binding=--interleave=2,3 tu-c0r2n78 - 3 device=3 binding=--interleave=6,7 tu-c0r2n93 - 3 device=3 binding=--interleave=6,7 tu-c0r2n78 - 2 device=2 binding=--interleave=4,5 tu-c0r2n93 - 1 device=1 binding=--interleave=2,3 tu-c0r2n93 - 0 device=0 binding=--interleave=0,1 tu-c0r2n93 - 2 device=2 binding=--interleave=4,5 tu-c0r2n90 - 2 device=2 binding=--interleave=4,5 tu-c0r2n87 - 1 device=1 binding=--interleave=2,3 tu-c0r2n87 - 2 device=2 binding=--interleave=4,5 tu-c0r2n87 - 0 device=0 binding=--interleave=0,1 tu-c0r2n87 - 3 device=3 binding=--interleave=6,7 tu-c0r2n90 - 1 device=1 binding=--interleave=2,3 tu-c0r2n90 - 0 device=0 binding=--interleave=0,1 tu-c0r2n90 - 3 device=3 binding=--interleave=6,7 tu-c0r2n84 - 3 device=3 binding=--interleave=6,7 tu-c0r2n84 - 0 device=0 binding=--interleave=0,1 tu-c0r2n84 - 1 device=1 binding=--interleave=2,3 tu-c0r2n84 - 2 device=2 binding=--interleave=4,5 tu-c0r2n81 - 3 device=3 binding=--interleave=6,7 tu-c0r2n81 - 0 device=0 binding=--interleave=0,1 tu-c0r2n81 - 2 device=2 binding=--interleave=4,5 tu-c0r2n81 - 1 device=1 binding=--interleave=2,3 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: 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 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 AcceleratorCudaInit: ================================================ local rank 0 device 0 bus id: 0000:03:00.0 AcceleratorCudaInit: ================================================ AcceleratorCudaInit: ================================================ AcceleratorCudaInit: ================================================ local rank 2 device 0 bus id: 0000:84:00.0 SharedMemoryMpi: World communicator of size 32 SharedMemoryMpi: Node communicator of size 4 0SharedMemoryMpi: SharedMemoryMPI.cc acceleratorAllocDevice 2147483648bytes at 0x152120000000 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 : 1.468074 s : Grid Layout Grid : Message : 1.468078 s : Global lattice size : 48 48 48 48 Grid : Message : 1.468084 s : OpenMP threads : 4 Grid : Message : 1.468087 s : MPI tasks : 2 2 2 4 Grid : Message : 1.480644 s : Making s innermost grids Grid : Message : 1.490829 s : Initialising 4d RNG Grid : Message : 1.507023 s : Intialising parallel RNG with unique string 'The 4D RNG' Grid : Message : 1.507053 s : Seed SHA256: 49db4542db694e3b1a74bf2592a8c1b83bfebbe18401693c2609a4c3af1 Grid : Message : 1.633148 s : Initialising 5d RNG Grid : Message : 1.867353 s : Intialising parallel RNG with unique string 'The 5D RNG' Grid : Message : 1.867389 s : Seed SHA256: b6316f2fac44ce14111f93e0296389330b077bfd0a7b359f781c58589f8a Grid : Message : 4.478085 s : Initialised RNGs Grid : Message : 5.492795 s : Drawing gauge field Grid : Message : 5.615970 s : Random gauge initialised Grid : Message : 5.620536 s : Setting up Cshift based reference Grid : Message : 10.664337 s : ***************************************************************** Grid : Message : 10.664358 s : * Kernel options --dslash-generic, --dslash-unroll, --dslash-asm Grid : Message : 10.664359 s : ***************************************************************** Grid : Message : 10.664360 s : ***************************************************************** Grid : Message : 10.664361 s : * Benchmarking DomainWallFermionR::Dhop Grid : Message : 10.664362 s : * Vectorising space-time by 8 Grid : Message : 10.664363 s : * VComplexF size is 64 B Grid : Message : 10.664364 s : * SINGLE precision Grid : Message : 10.664365 s : * Using Overlapped Comms/Compute Grid : Message : 10.664366 s : * Using GENERIC Nc WilsonKernels Grid : Message : 10.664367 s : ***************************************************************** Grid : Message : 11.214611 s : Called warmup Grid : Message : 98.363620 s : Called Dw 30000 times in 8.71488e+07 us Grid : Message : 98.363662 s : mflop/s = 3.85939e+07 Grid : Message : 98.363664 s : mflop/s per rank = 1.20606e+06 Grid : Message : 98.363666 s : mflop/s per node = 4.82424e+06 Grid : Message : 98.363668 s : RF GiB/s (base 2) = 78421.9 Grid : Message : 98.363670 s : mem GiB/s (base 2) = 49013.7 Grid : Message : 98.364241 s : norm diff 1.05759e-13 Grid : Message : 98.374195 s : #### Dhop calls report Grid : Message : 98.374203 s : WilsonFermion5D Number of DhopEO Calls : 60002 Grid : Message : 98.374207 s : WilsonFermion5D TotalTime /Calls : 1453.89 us Grid : Message : 98.374209 s : WilsonFermion5D CommTime /Calls : 1000.15 us Grid : Message : 98.374211 s : WilsonFermion5D FaceTime /Calls : 233.257 us Grid : Message : 98.374213 s : WilsonFermion5D ComputeTime1/Calls : 2.80331 us Grid : Message : 98.374215 s : WilsonFermion5D ComputeTime2/Calls : 234.002 us Grid : Message : 98.374240 s : Average mflops/s per call : 1.77267e+10 Grid : Message : 98.374244 s : Average mflops/s per call per rank : 5.53959e+08 Grid : Message : 98.374246 s : Average mflops/s per call per node : 2.21584e+09 Grid : Message : 98.374248 s : Average mflops/s per call (full) : 3.92575e+07 Grid : Message : 98.374250 s : Average mflops/s per call per rank (full): 1.2268e+06 Grid : Message : 98.374252 s : Average mflops/s per call per node (full): 4.90719e+06 Grid : Message : 98.374254 s : WilsonFermion5D Stencil Grid : Message : 98.374255 s : WilsonFermion5D StencilEven Grid : Message : 98.374256 s : WilsonFermion5D StencilOdd Grid : Message : 98.374257 s : WilsonFermion5D Stencil Reporti() Grid : Message : 98.374259 s : WilsonFermion5D StencilEven Reporti() Grid : Message : 98.374260 s : WilsonFermion5D StencilOdd Reporti() Grid : Message : 107.960590 s : Compare to naive wilson implementation Dag to verify correctness Grid : Message : 107.960990 s : Called DwDag Grid : Message : 107.961010 s : norm dag result 12.0422 Grid : Message : 107.101567 s : norm dag ref 12.0422 Grid : Message : 107.104533 s : norm dag diff 7.13141e-14 Grid : Message : 107.117796 s : Calling Deo and Doe and //assert Deo+Doe == Dunprec Grid : Message : 107.179312 s : src_e0.499992 Grid : Message : 107.253994 s : src_o0.500008 Grid : Message : 107.270696 s : ********************************************************* Grid : Message : 107.270705 s : * Benchmarking DomainWallFermionF::DhopEO Grid : Message : 107.270709 s : * Vectorising space-time by 8 Grid : Message : 107.270712 s : * SINGLE precision Grid : Message : 107.270715 s : * Using Overlapped Comms/Compute Grid : Message : 107.270716 s : * Using GENERIC Nc WilsonKernels Grid : Message : 107.270719 s : ********************************************************* Grid : Message : 154.135001 s : Deo mflop/s = 3.58924e+07 Grid : Message : 154.135030 s : Deo mflop/s per rank 1.12164e+06 Grid : Message : 154.135032 s : Deo mflop/s per node 4.48654e+06 Grid : Message : 154.135035 s : #### Dhop calls report Grid : Message : 154.135037 s : WilsonFermion5D Number of DhopEO Calls : 30001 Grid : Message : 154.135039 s : WilsonFermion5D TotalTime /Calls : 1561.97 us Grid : Message : 154.135041 s : WilsonFermion5D CommTime /Calls : 1050.89 us Grid : Message : 154.135043 s : WilsonFermion5D FaceTime /Calls : 298.978 us Grid : Message : 154.135045 s : WilsonFermion5D ComputeTime1/Calls : 4.84867 us Grid : Message : 154.135047 s : WilsonFermion5D ComputeTime2/Calls : 237.085 us Grid : Message : 154.135069 s : Average mflops/s per call : 1.02257e+10 Grid : Message : 154.135073 s : Average mflops/s per call per rank : 3.19554e+08 Grid : Message : 154.135075 s : Average mflops/s per call per node : 1.27822e+09 Grid : Message : 154.135077 s : Average mflops/s per call (full) : 3.65411e+07 Grid : Message : 154.135083 s : Average mflops/s per call per rank (full): 1.14191e+06 Grid : Message : 154.135086 s : Average mflops/s per call per node (full): 4.56764e+06 Grid : Message : 154.135088 s : WilsonFermion5D Stencil Grid : Message : 154.135089 s : WilsonFermion5D StencilEven Grid : Message : 154.135092 s : WilsonFermion5D StencilOdd Grid : Message : 154.135095 s : WilsonFermion5D Stencil Reporti() Grid : Message : 154.135098 s : WilsonFermion5D StencilEven Reporti() Grid : Message : 154.135100 s : WilsonFermion5D StencilOdd Reporti() Grid : Message : 154.154317 s : r_e6.02129 Grid : Message : 154.156162 s : r_o6.02097 Grid : Message : 154.157526 s : res12.0423 Grid : Message : 154.263040 s : norm diff 0 Grid : Message : 154.406823 s : norm diff even 0 Grid : Message : 154.478221 s : norm diff odd 0