1
0
mirror of https://github.com/paboyle/Grid.git synced 2024-09-20 09:15:38 +01:00
Grid/systems/Perlmutter/dwf.64.64.64.64.4node.opt0

157 lines
9.5 KiB
Plaintext
Raw Normal View History

2021-09-22 14:02:34 +01:00
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 0x7fd460000000 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 : 0.667601 s : Grid Layout
Grid : Message : 0.667602 s : Global lattice size : 64 64 64 64
Grid : Message : 0.667610 s : OpenMP threads : 32
Grid : Message : 0.667611 s : MPI tasks : 2 2 2 2
Grid : Message : 0.702872 s : Making s innermost grids
Grid : Message : 0.742911 s : Initialising 4d RNG
Grid : Message : 0.813463 s : Intialising parallel RNG with unique string 'The 4D RNG'
Grid : Message : 0.813479 s : Seed SHA256: 49db4542db694e3b1a74bf2592a8c1b83bfebbe18401693c2609a4c3af1
Grid : Message : 0.922630 s : Initialising 5d RNG
Grid : Message : 2.306290 s : Intialising parallel RNG with unique string 'The 5D RNG'
Grid : Message : 2.306540 s : Seed SHA256: b6316f2fac44ce14111f93e0296389330b077bfd0a7b359f781c58589f8a
Grid : Message : 3.878430 s : Initialised RNGs
Grid : Message : 4.536926 s : Drawing gauge field
Grid : Message : 4.824391 s : Random gauge initialised
Grid : Message : 6.253195 s : Setting up Cshift based reference
Grid : Message : 7.326402 s : *****************************************************************
Grid : Message : 7.326411 s : * Kernel options --dslash-generic, --dslash-unroll, --dslash-asm
Grid : Message : 7.326412 s : *****************************************************************
Grid : Message : 7.326412 s : *****************************************************************
Grid : Message : 7.326412 s : * Benchmarking DomainWallFermionR::Dhop
Grid : Message : 7.326412 s : * Vectorising space-time by 8
Grid : Message : 7.326413 s : * VComplexF size is 64 B
Grid : Message : 7.326414 s : * SINGLE precision
Grid : Message : 7.326414 s : * Using Overlapped Comms/Compute
Grid : Message : 7.326414 s : * Using GENERIC Nc WilsonKernels
Grid : Message : 7.326414 s : *****************************************************************
Grid : Message : 8.283417 s : Called warmup
Grid : Message : 89.658859 s : Called Dw 3000 times in 8.13753e+07 us
Grid : Message : 89.658898 s : mflop/s = 1.3063e+07
Grid : Message : 89.658900 s : mflop/s per rank = 816437
Grid : Message : 89.658901 s : mflop/s per node = 3.26575e+06
Grid : Message : 89.658902 s : RF GiB/s (base 2) = 26543.7
Grid : Message : 89.658903 s : mem GiB/s (base 2) = 16589.8
Grid : Message : 89.662424 s : norm diff 1.03481e-13
Grid : Message : 89.700433 s : #### Dhop calls report
Grid : Message : 89.700452 s : WilsonFermion5D Number of DhopEO Calls : 6002
Grid : Message : 89.700456 s : WilsonFermion5D TotalTime /Calls : 13588.2 us
Grid : Message : 89.700457 s : WilsonFermion5D CommTime /Calls : 12137.3 us
Grid : Message : 89.700458 s : WilsonFermion5D FaceTime /Calls : 548.408 us
Grid : Message : 89.700459 s : WilsonFermion5D ComputeTime1/Calls : 42.6163 us
Grid : Message : 89.700460 s : WilsonFermion5D ComputeTime2/Calls : 910.312 us
Grid : Message : 89.700477 s : Average mflops/s per call : 4.43502e+09
Grid : Message : 89.700493 s : Average mflops/s per call per rank : 2.77189e+08
Grid : Message : 89.700494 s : Average mflops/s per call per node : 1.10875e+09
Grid : Message : 89.700495 s : Average mflops/s per call (full) : 1.32753e+07
Grid : Message : 89.700496 s : Average mflops/s per call per rank (full): 829709
Grid : Message : 89.700497 s : Average mflops/s per call per node (full): 3.31884e+06
Grid : Message : 89.700498 s : WilsonFermion5D Stencil
Grid : Message : 89.700498 s : WilsonFermion5D StencilEven
Grid : Message : 89.700498 s : WilsonFermion5D StencilOdd
Grid : Message : 89.700499 s : WilsonFermion5D Stencil Reporti()
Grid : Message : 89.700499 s : WilsonFermion5D StencilEven Reporti()
Grid : Message : 89.700499 s : WilsonFermion5D StencilOdd Reporti()
Grid : Message : 101.462401 s : Compare to naive wilson implementation Dag to verify correctness
Grid : Message : 101.462412 s : Called DwDag
Grid : Message : 101.462413 s : norm dag result 12.0421
Grid : Message : 101.474097 s : norm dag ref 12.0421
Grid : Message : 101.489396 s : norm dag diff 7.63236e-14
Grid : Message : 101.529094 s : Calling Deo and Doe and //assert Deo+Doe == Dunprec
Grid : Message : 101.996820 s : src_e0.499997
Grid : Message : 102.626690 s : src_o0.500003
Grid : Message : 102.125734 s : *********************************************************
Grid : Message : 102.125736 s : * Benchmarking DomainWallFermionF::DhopEO
Grid : Message : 102.125737 s : * Vectorising space-time by 8
Grid : Message : 102.125738 s : * SINGLE precision
Grid : Message : 102.125739 s : * Using Overlapped Comms/Compute
Grid : Message : 102.125739 s : * Using GENERIC Nc WilsonKernels
Grid : Message : 102.125739 s : *********************************************************
Grid : Message : 143.296910 s : Deo mflop/s = 1.30119e+07
Grid : Message : 143.297140 s : Deo mflop/s per rank 813244
Grid : Message : 143.297160 s : Deo mflop/s per node 3.25297e+06
Grid : Message : 143.297180 s : #### Dhop calls report
Grid : Message : 143.297190 s : WilsonFermion5D Number of DhopEO Calls : 3001
Grid : Message : 143.297200 s : WilsonFermion5D TotalTime /Calls : 13630 us
Grid : Message : 143.297210 s : WilsonFermion5D CommTime /Calls : 12124.9 us
Grid : Message : 143.297220 s : WilsonFermion5D FaceTime /Calls : 590.958 us
Grid : Message : 143.297230 s : WilsonFermion5D ComputeTime1/Calls : 43.2806 us
Grid : Message : 143.297240 s : WilsonFermion5D ComputeTime2/Calls : 921.187 us
Grid : Message : 143.297460 s : Average mflops/s per call : 4.24329e+09
Grid : Message : 143.297470 s : Average mflops/s per call per rank : 2.65206e+08
Grid : Message : 143.297480 s : Average mflops/s per call per node : 1.06082e+09
Grid : Message : 143.297490 s : Average mflops/s per call (full) : 1.32347e+07
Grid : Message : 143.297500 s : Average mflops/s per call per rank (full): 827169
Grid : Message : 143.297510 s : Average mflops/s per call per node (full): 3.30868e+06
Grid : Message : 143.297520 s : WilsonFermion5D Stencil
Grid : Message : 143.297520 s : WilsonFermion5D StencilEven
Grid : Message : 143.297520 s : WilsonFermion5D StencilOdd
Grid : Message : 143.297520 s : WilsonFermion5D Stencil Reporti()
Grid : Message : 143.297520 s : WilsonFermion5D StencilEven Reporti()
Grid : Message : 143.297520 s : WilsonFermion5D StencilOdd Reporti()
Grid : Message : 143.112368 s : r_e6.02111
Grid : Message : 143.119760 s : r_o6.02102
Grid : Message : 143.126239 s : res12.0421
Grid : Message : 143.720780 s : norm diff 0
Grid : Message : 144.885380 s : norm diff even 0
Grid : Message : 144.154396 s : norm diff odd 0