portelli/Grid
1
0
Fork 0
mirror of https://github.com/paboyle/Grid.git synced 2026-04-09 05:26:10 +01:00
Code Releases Activity

Commit Graph

  • 4b24800132 AVX512 drop mixed precision as well Luchang Jin 2021-09-15 16:24:01 -04:00
  • 9d2238148c Merge branch 'develop' of https://www.github.com/paboyle/Grid into develop Peter Boyle 2021-09-15 19:25:57 +01:00
  • c15493218d Two extra routines to break out SchurRedBlack on many RHS into stages to allow efficient deflation & split grid Split grid solver still to do. Peter Boyle 2021-09-15 19:24:39 +01:00
  • 001a556a34 Merge pull request #365 from lehner/feature/gpt Peter Boyle 2021-09-15 13:34:02 -04:00
  • 3d0f88e702 A64FX drop mixed precision as well Christoph Lehner 2021-09-15 18:38:32 +02:00
  • dd091d0960 consistent pointer offloading instead of views Christoph Lehner 2021-09-15 16:58:05 +02:00
  • e2abbf9520 Merge pull request #25 from paboyle/develop Christoph Lehner 2021-09-15 10:02:43 +02:00
  • c7baeb5bae Merge branch 'develop' of https://github.com/paboyle/Grid into develop Peter Boyle 2021-09-14 08:31:11 -07:00
  • 402d80e197 Merge branch 'develop' of https://www.github.com/paboyle/Grid into develop Peter Boyle 2021-09-14 16:16:06 +01:00
  • 86e33c8ab2 Significant GPU perf speed up finished Peter Boyle 2021-09-14 16:14:23 +01:00
  • 5dae6a6dac Deprecate half prec comms Peter Boyle 2021-09-14 15:06:59 +01:00
  • 361bb8a101 Remove half prec comms Peter Boyle 2021-09-14 15:06:29 +01:00
  • 7efdb3cd2b Remove half prec comms Peter Boyle 2021-09-14 15:06:06 +01:00
  • 65ef4ec29f Move tables to device memory Peter Boyle 2021-09-14 15:05:01 +01:00
  • d5835c0222 Switch to coalesced stencil face gather Peter Boyle 2021-09-14 15:04:14 +01:00
  • a7b943b33e Remove half prec comms Peter Boyle 2021-09-14 05:05:33 +01:00
  • 7440cde92f No half prec comms; coalesced access on GPU Peter Boyle 2021-09-14 05:04:56 +01:00
  • 0fc662bb24 Dirac cuda 11.4 happy ; force host for functions accessing mult table ET runs these on host BEFORE lodging result in AST for kernel Peter Boyle 2021-09-14 05:00:44 +01:00
  • 8195890640 Force MPI over NVLINK Peter Boyle 2021-09-14 05:00:17 +01:00
  • 4c88104a73 Fix compile warns Peter Boyle 2021-09-11 23:08:05 +01:00
  • 73b944c152 Drop half prec comms for now. Peter Boyle 2021-09-11 23:07:18 +01:00
  • d1b0b7f5c6 Half prec comms dropping Peter Boyle 2021-09-11 23:05:40 +01:00
  • 381d8797d0 Drop half prec comms for now Peter Boyle 2021-09-11 23:05:02 +01:00
  • d184b8c921 Merge branch 'develop' into gparity_HMC Christopher Kelly 2021-09-08 06:14:08 -07:00
  • c92e390b08 Added initial main binary code for 40ID and 48ID Gparity HMC Christopher Kelly 2021-09-08 09:00:13 -04:00
  • 8fd16686dc Checking block lanczos deleted some diag outputs Chulwoo Jung 2021-09-05 23:04:41 -04:00
  • 23b9c6b5f5 Merge branch 'develop' of https://github.com/paboyle/Grid into feature/block_lanczos Chulwoo Jung 2021-09-03 17:38:10 -04:00
  • 11ee8a1061 Merge remote-tracking branch 'upstream/develop' into develop Henrique Rocha 2021-09-02 16:57:42 +01:00
  • b06526bc1e Comment update Peter Boyle 2021-08-30 21:15:39 -04:00
  • 3044419111 Some sample code Peter Boyle 2021-08-30 20:32:11 -04:00
  • bcfa9cf068 Improvement of output Peter Boyle 2021-08-28 08:08:15 -07:00
  • 114920b8de Some example clean up Peter Boyle 2021-08-25 12:24:17 +01:00
  • 0d588b95f4 Bug fix to Example_Laplacian test Peter Boyle 2021-08-23 23:14:26 +01:00
  • 5b3c530aa7 Return value Peter Boyle 2021-08-23 15:30:45 +01:00
  • c6a5499c8b Fail on non-apple Peter Boyle 2021-08-22 18:40:55 +01:00
  • ec9c3fe77a Remove the file Peter Boyle 2021-08-22 18:28:39 +01:00
  • 6135ad530e Extra examples / solutions Peter Boyle 2021-08-22 18:25:07 +01:00
  • 40098424c7 Examples Peter Boyle 2021-08-22 14:17:12 +01:00
  • 7163b31a26 Examples Peter Boyle 2021-08-20 01:15:23 +01:00
  • ffbdd91e0e Apple happiness Peter Boyle 2021-08-20 01:15:00 +01:00
  • 7760417312 Production jobs Peter Boyle 2021-08-12 00:18:34 +02:00
  • 5d29e175d8 Typo fix Peter Boyle 2021-08-10 18:25:43 +01:00
  • 417dbfa257 Fix Peter Boyle 2021-08-10 08:55:35 -07:00
  • 1eda4d8e0b Merge branch 'develop' of https://github.com/paboyle/Grid into develop peterx.a.boyle 2021-08-10 05:41:18 -07:00
  • 50181f16e5 Level 0 IPC set up peterx.a.boyle 2021-08-10 05:35:15 -07:00
  • 75030637cc Improved comms benchmark, same as benchmark_comms_host_device Peter Boyle 2021-08-10 05:16:30 -07:00
  • fe5aaf7677 Make comms benchmark same as Benchmark_comms_host_device Peter Boyle 2021-08-09 04:06:30 -07:00
  • 80ac2a73ca Check is wrong (HtoD / DtoH) Peter Boyle 2021-08-05 18:33:20 -04:00
  • 770680669d Whitespace removal. Andrew Yong 2021-08-04 09:21:59 +01:00
  • 0cdfc5cf22 Merge remote-tracking branch 'upstream/develop' into develop Andrew Yong 2021-07-30 14:40:55 +01:00
  • 5b36a8af54 Added a CshiftLink function to the GaugeImplementations and boundary condition classes that offers a boundary aware C-shift Modified gauge fixing code to use CshiftLink internally such that the steepest descent algorithm is universal Modified gauge transformation code to use CshiftLink for a universal definition Improved comprehensibility of Test_fft_gfix and generalized to use either periodic or charge conjugation BCs based on cmdline option Added cmdline options to Test_fft_gfix to tune alpha and optionally disable the Fourier acceleration tests Christopher Kelly 2021-07-12 17:13:40 -04:00
  • d75a66a3e6 test done Felix Erben 2021-07-06 11:42:36 +01:00
  • fcc4374d7b i/o done Felix Erben 2021-07-05 14:52:00 +01:00
  • 67c3c16fe5 working test Felix Erben 2021-07-05 14:41:52 +01:00
  • 25e9be50b5 created test file Felix Erben 2021-07-02 15:51:19 +01:00
  • 75a1f85162 Added method to compute and return the Wilson flow energy density over some number of steps Christopher Kelly 2021-06-30 17:24:00 -04:00
  • 428b8ba907 Updated from upstream and added halo benchmark Henrique B.R 2021-06-29 01:05:12 +01:00
  • 323cf6c038 make message consistent with configure script Ed Bennett 2021-06-23 17:00:43 +01:00
  • 29a22ae603 Simpler SYCL setup Peter Boyle 2021-06-22 17:57:20 +00:00
  • 403bff1a47 Force reqd subgroup size fo SYCL Peter Boyle 2021-06-22 17:56:10 +00:00
  • c50f27e68b Make FFT play nice with split grid Christoph Lehner 2021-06-20 11:34:38 +02:00
  • 80afacec5b nvcc: Add -fopenmp to LDFLAGS Peter Georg 2021-06-17 13:05:13 +02:00
  • d4290a7434 finer timers in Benchmark_IO feature/benchiotimings Antonin Portelli 2021-06-17 11:57:02 +01:00
  • 6cd9224dd7 SYCL comms buffer allocate Peter Boyle 2021-06-16 17:10:55 +00:00
  • 4bf8196ff1 Merge branch 'develop' of https://www.github.com/paboyle/Grid into develop Peter Boyle 2021-06-15 21:45:36 +00:00
  • 4c5440fb06 const happy for sycl Peter Boyle 2021-06-15 21:45:07 +00:00
  • 8cc0defb69 Reorg headers Quadro 2021-06-10 14:28:49 -04:00
  • e0c9d01123 Build tests works Quadro 2021-06-10 14:28:26 -04:00
  • 91bf1df018 Rename changes Quadro 2021-06-09 22:58:01 -04:00
  • 386a89c668 Updated mixed prec Quadro 2021-06-09 17:14:24 -04:00
  • 670f4985fd DDHMC test update Quadro 2021-06-09 16:35:53 -04:00
  • dcd48a0435 Domain decomposed benchmark Quadro 2021-06-09 16:35:24 -04:00
  • 87ec14c353 Update for mixed precison solve Quadro 2021-06-09 16:34:44 -04:00
  • 4f5ad73501 Mixed prec update Quadro 2021-06-09 16:33:02 -04:00
  • 096bb98f78 Remove Quadro 2021-06-09 16:32:35 -04:00
  • 76837ffc65 Working (I think) version Quadro 2021-06-09 16:31:37 -04:00
  • 81bd0d7906 Default to mixed precision now Quadro 2021-06-09 16:31:19 -04:00
  • 7d8d250389 Complete ? Quadro 2021-06-09 16:30:39 -04:00
  • ad406b2c3e Provide a call back for subdomain solve / local domain pseudofermion restriction Quadro 2021-06-09 16:29:15 -04:00
  • e6366b98a5 Mixed precision & domain changes Quadro 2021-06-09 16:27:59 -04:00
  • 302356189c Reduce verbose Quadro 2021-06-09 14:56:49 -04:00
  • 9394450c1a Verbose changes Quadro 2021-06-09 13:30:42 -04:00
  • 6cf3edef00 More logging Quadro 2021-06-09 12:59:27 -04:00
  • 31cc227dd2 Domain definition class Quadro 2021-06-09 12:58:57 -04:00
  • c690e66325 Mobius two flavour Quadro 2021-06-09 12:58:24 -04:00
  • 5fdbb924f0 Mixed prec tolerance changes Quadro 2021-06-09 12:57:03 -04:00
  • 6dcaed621c General detection for arbitrary domains. Can simplify and make specific if performance matters Quadro 2021-06-09 12:54:43 -04:00
  • ac4f2d9798 Fixed EOFA approx test square rooting the result inappropriately thus failing when it shouldn't To MDWF+ID GPBC evol main program, added routine to compute the lower bound of the EOFA using the power method with a command line toggle Christopher Kelly 2021-06-09 09:08:37 -04:00
  • a269a3d919 Merge pull request #358 from mmphys/feature/serialisation-test feature/serialisation-update Antonin Portelli 2021-06-09 10:16:25 +01:00
  • f9cda24781 Cleaner Quadro 2021-06-08 21:50:43 -04:00
  • cd5e3fbd82 berrier for debug code Quadro 2021-06-08 21:18:22 -04:00
  • f68036c79f Inner mixed tolerance Quadro 2021-06-08 21:17:06 -04:00
  • 216c178c16 Clean up Quadro 2021-06-08 20:59:00 -04:00
  • 990d976241 More precision to convince Chris Kelly it's right Quadro 2021-06-08 20:29:27 -04:00
  • f2fe2573a7 Constructor control of inner tolerance Quadro 2021-06-08 20:27:52 -04:00
  • 11c55a0476 More verbose Quadro 2021-06-08 20:26:52 -04:00
  • 0c4f585496 Test nested std::vector<grid tensor> Michael Marshall 2021-06-08 00:05:35 +01:00
  • 33d2df46a0 Merge branch 'develop' into feature/serialisation-test Michael Marshall 2021-06-07 23:25:38 +01:00
  • 2df308f649 Add a ragged vector to the serialisation tests. NB: Already had nested (regular) std::vector<std::vector<...>> Michael Marshall 2021-06-07 23:25:07 +01:00
  • b284d50863 Checking in fixed adaptive WilsonFlow feature/adaptive_wflow Chulwoo Jung 2021-06-07 14:20:27 -04:00