portelli/Grid
1
0
Fork 0
mirror of https://github.com/paboyle/Grid.git synced 2024-11-15 02:05:37 +00:00
Code Releases Activity
655c79f39e
Grid/systems/Booster/dwf.16node.perf

157 lines
9.6 KiB
Plaintext
Raw Normal View History

Udpdate
2021-09-21 00:04:30 +01:00
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 64
SharedMemoryMpi: Node communicator of size 4
0SharedMemoryMpi: SharedMemoryMPI.cc acceleratorAllocDevice 2147483648bytes at 0x14ac40000000 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=f660dc67e4b193afc4015bc5e5fe47cfdbb0356e: (HEAD -> develop, origin/develop, origin/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 32 LARGE 8
Grid : Message : MemoryManager::Init() Non unified: Caching accelerator data in dedicated memory
Grid : Message : MemoryManager::Init() Using cudaMalloc
Grid : Message : 0.910318 s : Grid Layout
Grid : Message : 0.910320 s : Global lattice size : 64 64 64 256
Grid : Message : 0.910325 s : OpenMP threads : 4
Grid : Message : 0.910326 s : MPI tasks : 2 2 2 8
Grid : Message : 0.973956 s : Making s innermost grids
Grid : Message : 1.198830 s : Initialising 4d RNG
Grid : Message : 1.119813 s : Intialising parallel RNG with unique string 'The 4D RNG'
Grid : Message : 1.119870 s : Seed SHA256: 49db4542db694e3b1a74bf2592a8c1b83bfebbe18401693c2609a4c3af1
Grid : Message : 2.683307 s : Initialising 5d RNG
Grid : Message : 4.220535 s : Intialising parallel RNG with unique string 'The 5D RNG'
Grid : Message : 4.220563 s : Seed SHA256: b6316f2fac44ce14111f93e0296389330b077bfd0a7b359f781c58589f8a
Grid : Message : 37.198140 s : Initialised RNGs
Grid : Message : 39.952612 s : Drawing gauge field
Grid : Message : 40.488019 s : Random gauge initialised
Grid : Message : 42.659220 s : Setting up Cshift based reference
Grid : Message : 47.622210 s : *****************************************************************
Grid : Message : 47.622236 s : * Kernel options --dslash-generic, --dslash-unroll, --dslash-asm
Grid : Message : 47.622237 s : *****************************************************************
Grid : Message : 47.622238 s : *****************************************************************
Grid : Message : 47.622239 s : * Benchmarking DomainWallFermionR::Dhop
Grid : Message : 47.622240 s : * Vectorising space-time by 8
Grid : Message : 47.622241 s : * VComplexF size is 64 B
Grid : Message : 47.622242 s : * SINGLE precision
Grid : Message : 47.622243 s : * Using Overlapped Comms/Compute
Grid : Message : 47.622244 s : * Using GENERIC Nc WilsonKernels
Grid : Message : 47.622245 s : *****************************************************************
Grid : Message : 48.950210 s : Called warmup
Grid : Message : 77.311124 s : Called Dw 3000 times in 2.83592e+07 us
Grid : Message : 77.311181 s : mflop/s = 1.49934e+08
Grid : Message : 77.311184 s : mflop/s per rank = 2.34273e+06
Grid : Message : 77.311185 s : mflop/s per node = 9.37091e+06
Grid : Message : 77.311186 s : RF GiB/s (base 2) = 304663
Grid : Message : 77.311187 s : mem GiB/s (base 2) = 190415
Grid : Message : 77.314752 s : norm diff 1.03478e-13
Grid : Message : 77.349587 s : #### Dhop calls report
Grid : Message : 77.349591 s : WilsonFermion5D Number of DhopEO Calls : 6002
Grid : Message : 77.349613 s : WilsonFermion5D TotalTime /Calls : 4761.53 us
Grid : Message : 77.349615 s : WilsonFermion5D CommTime /Calls : 3363.09 us
Grid : Message : 77.349616 s : WilsonFermion5D FaceTime /Calls : 469.094 us
Grid : Message : 77.349617 s : WilsonFermion5D ComputeTime1/Calls : 26.8794 us
Grid : Message : 77.349618 s : WilsonFermion5D ComputeTime2/Calls : 949.276 us
Grid : Message : 77.349702 s : Average mflops/s per call : 2.68569e+10
Grid : Message : 77.349710 s : Average mflops/s per call per rank : 4.1964e+08
Grid : Message : 77.349711 s : Average mflops/s per call per node : 1.67856e+09
Grid : Message : 77.349712 s : Average mflops/s per call (full) : 1.51538e+08
Grid : Message : 77.349713 s : Average mflops/s per call per rank (full): 2.36779e+06
Grid : Message : 77.349714 s : Average mflops/s per call per node (full): 9.47115e+06
Grid : Message : 77.349715 s : WilsonFermion5D Stencil
Grid : Message : 77.349716 s : WilsonFermion5D StencilEven
Grid : Message : 77.349717 s : WilsonFermion5D StencilOdd
Grid : Message : 77.349718 s : WilsonFermion5D Stencil Reporti()
Grid : Message : 77.349719 s : WilsonFermion5D StencilEven Reporti()
Grid : Message : 77.349720 s : WilsonFermion5D StencilOdd Reporti()
Grid : Message : 104.883719 s : Compare to naive wilson implementation Dag to verify correctness
Grid : Message : 104.883743 s : Called DwDag
Grid : Message : 104.883744 s : norm dag result 12.0421
Grid : Message : 104.901901 s : norm dag ref 12.0421
Grid : Message : 104.917822 s : norm dag diff 7.63254e-14
Grid : Message : 104.957229 s : Calling Deo and Doe and //assert Deo+Doe == Dunprec
Grid : Message : 105.334551 s : src_e0.499998
Grid : Message : 105.416616 s : src_o0.500002
Grid : Message : 105.486729 s : *********************************************************
Grid : Message : 105.486732 s : * Benchmarking DomainWallFermionF::DhopEO
Grid : Message : 105.486733 s : * Vectorising space-time by 8
Grid : Message : 105.486734 s : * SINGLE precision
Grid : Message : 105.486739 s : * Using Overlapped Comms/Compute
Grid : Message : 105.486740 s : * Using GENERIC Nc WilsonKernels
Grid : Message : 105.486741 s : *********************************************************
Grid : Message : 119.695464 s : Deo mflop/s = 1.5039e+08
Grid : Message : 119.695494 s : Deo mflop/s per rank 2.34984e+06
Grid : Message : 119.695496 s : Deo mflop/s per node 9.39937e+06
Grid : Message : 119.695502 s : #### Dhop calls report
Grid : Message : 119.695503 s : WilsonFermion5D Number of DhopEO Calls : 3001
Grid : Message : 119.695505 s : WilsonFermion5D TotalTime /Calls : 4734.45 us
Grid : Message : 119.695507 s : WilsonFermion5D CommTime /Calls : 3287.23 us
Grid : Message : 119.695508 s : WilsonFermion5D FaceTime /Calls : 537.724 us
Grid : Message : 119.695509 s : WilsonFermion5D ComputeTime1/Calls : 16.0483 us
Grid : Message : 119.695510 s : WilsonFermion5D ComputeTime2/Calls : 939.854 us
Grid : Message : 119.695533 s : Average mflops/s per call : 4.50726e+10
Grid : Message : 119.695535 s : Average mflops/s per call per rank : 7.04259e+08
Grid : Message : 119.695536 s : Average mflops/s per call per node : 2.81703e+09
Grid : Message : 119.695537 s : Average mflops/s per call (full) : 1.52405e+08
Grid : Message : 119.695538 s : Average mflops/s per call per rank (full): 2.38133e+06
Grid : Message : 119.695539 s : Average mflops/s per call per node (full): 9.52532e+06
Grid : Message : 119.695540 s : WilsonFermion5D Stencil
Grid : Message : 119.695541 s : WilsonFermion5D StencilEven
Grid : Message : 119.695542 s : WilsonFermion5D StencilOdd
Grid : Message : 119.695543 s : WilsonFermion5D Stencil Reporti()
Grid : Message : 119.695544 s : WilsonFermion5D StencilEven Reporti()
Grid : Message : 119.695545 s : WilsonFermion5D StencilOdd Reporti()
Grid : Message : 119.752707 s : r_e6.02108
Grid : Message : 119.759448 s : r_o6.02101
Grid : Message : 119.765382 s : res12.0421
Grid : Message : 120.419093 s : norm diff 0
Grid : Message : 120.829772 s : norm diff even 0
Grid : Message : 120.909078 s : norm diff odd 0
Copy Permalink