portelli/Grid
1
0
Fork 0
mirror of https://github.com/paboyle/Grid.git synced 2025-04-24 12:45:56 +01:00
Code Releases Activity

Commit Graph

  • a957e7bfa1 Adding DWF evec Chirality measurement specflow Chulwoo Jung 2025-04-22 22:17:51 +00:00
  • cee4c8ce8c Merge branch 'develop' of https://github.com/paboyle/Grid into specflow Chulwoo Jung 2025-04-18 19:55:36 +00:00
  • e652fc2825 Shared Memory test reenabled on every Grid object creation. Const improvements in Accelerator.h develop Peter Boyle 2025-04-07 11:51:40 -04:00
  • a49fa3f8d0 ROCM 6.3.1 appears to work Peter Boyle 2025-04-07 11:50:59 -04:00
  • cd452a2f91 Slurm update Peter Boyle 2025-04-04 18:34:53 -04:00
  • 4f89f603ae Changes to add back shared memory test on GPU Peter Boyle 2025-04-04 18:34:36 -04:00
  • 11dc2c5e1d PVdagM initialise Peter Boyle 2025-04-04 16:48:36 -04:00
  • 6fec3c15ca Cleaner printing Peter Boyle 2025-04-04 16:46:57 -04:00
  • 938c47480f Updated compile on frontier. Unsatisfactory hacsk Peter Boyle 2025-04-04 16:46:27 -04:00
  • 3811d19298 Fence Peter Boyle 2025-04-04 16:45:32 -04:00
  • 83a3ab6b6f Barrier -- not sure 100% this was needed Peter Boyle 2025-04-04 16:44:40 -04:00
  • d66a9af6a3 No compile fix Peter Boyle 2025-04-04 16:44:12 -04:00
  • adc90d3a86 NVLINK GET/PUT on cuda aware mpi Peter Boyle 2025-04-04 16:43:36 -04:00
  • ebbd015c5c Deprecate shared memory copy as direction matters on nvidia GPU Peter Boyle 2025-04-04 16:42:15 -04:00
  • 4ab73b36b2 Deprecate shared memory copy as direction matters on GPU Peter Boyle 2025-04-04 16:41:37 -04:00
  • 130e07a422 Non hermitian support Peter Boyle 2025-04-04 16:39:58 -04:00
  • 8f47bb367e Shifted non herm Peter Boyle 2025-04-04 16:39:24 -04:00
  • 0c3cb60135 Script update Peter Boyle 2025-02-17 15:02:10 -05:00
  • 9eae8fca5d Size outut Peter Boyle 2025-02-17 15:08:25 -05:00
  • 882a217074 Example of Useful prerequisite installs with spack Peter Boyle 2025-03-26 11:27:59 -04:00
  • 199818bd6c
         Merge pull request #475 from lehner/feature-aurora Peter Boyle 2025-03-13 08:55:55 -04:00
  • fe66c7ca30 verbosity Christoph Lehner 2025-03-13 12:49:36 +00:00
  • e9177e4af3 Blas compatibility Christoph Lehner 2025-03-13 08:48:23 +00:00
  • d15a6c5933 Merge branch 'develop' of https://github.com/paboyle/Grid into feature-aurora Christoph Lehner 2025-03-13 07:29:55 +00:00
  • 25ab9325e7 Use hostVector but remove construct resize Peter Boyle 2025-03-11 15:02:32 +00:00
  • 19f9378b98 Should work on Aurora nowb Peter Boyle 2025-03-11 13:49:53 +00:00
  • 9ffd1ed4ce Merged Christoph Lehner 2025-03-08 15:30:08 +00:00
  • 3d014864e2 Makinig LLVM happy Peter Boyle 2025-03-06 14:18:43 -05:00
  • 1d22841811 Working on aurora, GPT issue turned up is fixed Peter Boyle 2025-03-06 03:19:35 +00:00
  • a1cdda833f
         Update WorkArounds.txt Peter Boyle 2025-03-05 14:04:23 -05:00
  • ad6db92690
         Update WorkArounds.txt Peter Boyle 2025-03-05 14:00:26 -05:00
  • e8ff9d8e50
         Update WorkArounds.txt Peter Boyle 2025-03-05 14:00:04 -05:00
  • 795769c636
         Update WorkArounds.txt Peter Boyle 2025-03-05 13:50:41 -05:00
  • 267a39d943
         Update WorkArounds.txt Peter Boyle 2025-03-05 13:49:43 -05:00
  • 3624bd3d22
         Update WorkArounds.txt Peter Boyle 2025-03-05 13:45:09 -05:00
  • bc12dbbb38
         Update WorkArounds.txt Peter Boyle 2025-03-05 12:48:56 -05:00
  • eb8a008a8f
         Create WorkArounds.txt Peter Boyle 2025-03-05 12:41:59 -05:00
  • c4d9aa1a21 Config command that makes GPT happier Peter Boyle 2025-02-27 20:12:12 +00:00
  • 6ae809ed40 Print not liked on GPT compile Peter Boyle 2025-02-27 20:11:45 +00:00
  • 311e2aab3f
         Update Accelerator.h Peter Boyle 2025-02-26 11:42:52 -05:00
  • 438dfbdb83 Only throw if there is a pending list entry in CommsComplete Peter Boyle 2025-02-25 16:57:27 +00:00
  • b2ce760cf4 Verbose issue with GPT Peter Boyle 2025-02-25 16:55:23 +00:00
  • ba9bbe0221 Bounce MPI through host Peter Boyle 2025-02-12 19:34:59 +00:00
  • 4c3dd82d84 CSHIFT with bounce throuhgh Host memory on MPI packets Peter Boyle 2025-02-12 19:09:53 +00:00
  • 44e911b5b7 Comment change Peter Boyle 2025-02-12 17:37:55 +00:00
  • a7a16df9d0 GET not put has kinder barrier sequence for NVLINK type access as when GET is done, I can use it without barrier. Moves a barrier to a nicer place, overlapped with DtoH DMA Peter Boyle 2025-02-12 14:59:28 +00:00
  • 382e0abefd Was issueing a double fence -- the gather also fences Peter Boyle 2025-02-12 14:57:28 +00:00
  • 6fdefe5b90 Barrier sequencing if doing "GET" not "PUT" is different. This is somewhat better timing for Barriers Peter Boyle 2025-02-12 14:55:20 +00:00
  • 4788dd8e2e More states in packet progression for GPU non aware MPI Peter Boyle 2025-02-12 14:53:57 +00:00
  • 1cc5f221f3 GET not put ordering is better as I know when I've got all MY data Peter Boyle 2025-02-12 14:53:05 +00:00
  • 93251bfba0 GET not put for better ordering in the downstream dependent kernels -- I know when I'm done, so we can move a barrier / handshake between ranks intranode to a point off critical path Peter Boyle 2025-02-12 14:50:21 +00:00
  • 18b79508b8 New line better for pretty print Peter Boyle 2025-02-12 14:49:48 +00:00
  • 4de5ed1613 Remove vector view. The std::vector will not inform Memory manager of deletion and so a stale entry could be left. It is not and should not be used. Peter Boyle 2025-02-12 14:48:46 +00:00
  • 0baaddbe98 Pipeline mode commit on Aurora. 5+ TF/s on 16^3x32 per tile at 384 nodes. More concurrency/fine grained scheduling is possible. Peter Boyle 2025-02-04 19:27:26 +00:00
  • b50fb34e71 Perf on Aurora Peter Boyle 2025-02-01 18:39:34 +00:00
  • de84d730ff Fastest run config on Aurora to date Peter Boyle 2025-02-01 18:07:42 +00:00
  • c74d11e3d7 PVdagM MG Peter Boyle 2025-02-01 11:04:13 -05:00
  • 84cab5e6e7 no comms and log cleanup Christoph Lehner 2025-02-01 16:37:21 +01:00
  • c4fc972fec Merge branch 'feature/deprecate-uvm' into develop Peter Boyle 2025-01-31 16:32:36 +00:00
  • 8cf809e231 Best results on Aurora so far feature/deprecate-uvm Peter Boyle 2025-01-31 16:14:45 +00:00
  • 94019a922e Significantly better performance on Aurora without using pipeline mode Peter Boyle 2025-01-30 16:36:46 +00:00
  • d6b2727f86 Pipeline mode getting better -- 2 nodes @ 10TF/s per node on Aurora Peter Boyle 2025-01-29 09:22:21 +00:00
  • 74a4f43946 Optional host buffer bounce for no CUDA aware MPI Peter Boyle 2025-01-28 15:22:46 +00:00
  • 1caf8b0f86 Rename Peter Boyle 2025-01-28 15:22:37 +00:00
  • 570b72a47b Bugfix. Sorry! Chulwoo Jung 2025-01-21 15:37:39 -05:00
  • a5798a89ed Merge branch 'develop' into specflow Chulwoo Jung 2025-01-21 12:13:24 -05:00
  • 3f3661a86f Heading towards PVdagM multigrid Peter Boyle 2025-01-17 14:33:35 +00:00
  • f7e2f9a401 Checking in spectral flow and DWF/Mobius kernel eigenvalue measurement Chulwoo Jung 2025-01-16 20:47:33 +00:00
  • 2848a9b558 DWF Kernel lanczos working(?) Chulwoo Jung 2025-01-16 01:29:56 +00:00
  • 8fe429346f Dslash testing for reproduce Peter Boyle 2024-11-11 23:11:11 +00:00
  • 5a4f9bf2e3 Force the ROCM version Peter Boyle 2024-10-29 18:12:31 -04:00
  • b91fc1b6b4 Merge branch 'feature/boosted' into feature/deprecate-uvm Peter Boyle 2024-10-28 16:53:09 -04:00
  • eafc150034 Test fft asserts Peter Boyle 2024-10-23 16:46:26 -04:00
  • 2877f1a268 Verbose reduce Peter Boyle 2024-10-23 15:14:16 -04:00
  • 1e893af775 GPU happy Peter Boyle 2024-10-23 14:52:15 -04:00
  • d9f430a575 Happy GPU Peter Boyle 2024-10-23 14:51:16 -04:00
  • 63abe87f36 Memory manager verbose improvements that were useful to track an error Peter Boyle 2024-10-23 14:49:13 -04:00
  • 368d649c8a feature/deprecate-uvm happier -- preallocate device resident neigbour table Peter Boyle 2024-10-23 14:47:55 -04:00
  • 5603464f39 Fix in partial fraction import/export physical and make the GPU happier on the deprecate-uvm -- don't use static vectors, make member of class Peter Boyle 2024-10-23 14:45:58 -04:00
  • 655c79f39e Suppress warning on partial override Peter Boyle 2024-10-23 14:44:41 -04:00
  • 565b231c03 Nvcc happy Peter Boyle 2024-10-23 14:44:17 -04:00
  • 62a9f180fa NVCC happy Peter Boyle 2024-10-23 14:44:04 -04:00
  • 5ae77876a8 Meson field and Aslash field on GPU; some compiler warning removed Peter Boyle 2024-10-18 19:08:06 -04:00
  • 4ed2c2c74f Config command Peter Boyle 2024-10-18 13:58:33 -04:00
  • 955da582b6 Working on NVCC Peter Boyle 2024-10-18 13:58:03 -04:00
  • 11b07b950d Vanilla linux compile, assuming spack prerequisites Peter Boyle 2024-10-18 13:57:40 -04:00
  • 8f70cfeda9 Clean up Peter Boyle 2024-10-18 13:56:53 -04:00
  • ce64271048 Remove the copying version Peter Boyle 2024-10-18 13:56:24 -04:00
  • 5cc4f3241d Meson field test Peter Boyle 2024-10-18 15:42:30 +00:00
  • 6815e138b4 Boosted fermion attempt feature/boosted Peter Boyle 2024-10-17 18:36:32 +01:00
  • a78a61d76f Update configure Peter Boyle 2024-10-15 14:38:45 +00:00
  • 2eff3f34ed Alternate reduction; default to grids own but make a configure flag --enable-reduction=grid|mpi Peter Boyle 2024-10-15 14:36:06 +00:00
  • 03687c1d62 Final version of test, closer to original again Peter Boyle 2024-10-15 14:35:17 +00:00
  • febfe4e77f Make my own reduction a configure flag Peter Boyle 2024-10-15 14:32:35 +00:00
  • 4d1aa134b5 Use normal reduction, configure flag to force deterministic Peter Boyle 2024-10-15 14:32:11 +00:00
  • 5ec879860a Odd rounding issue - bears looking into Peter Boyle 2024-10-15 14:30:54 +00:00
  • f617468e04
         Update Lattice_base.h Peter Boyle 2024-10-11 10:39:16 -04:00
  • b728af903c Fast axpy norm under CFLAG Peter Boyle 2024-10-11 03:23:09 +00:00
  • 54f1999030 axpy_norm_fast -- wasn't using the determinstic MPI sum causing issues Peter Boyle 2024-10-11 03:22:18 +00:00
  • fd58f0b669 Return ok Peter Boyle 2024-10-11 03:21:21 +00:00