Commit Graph

  • 2e558edcb9 [ci] Adding composablekernel to TheRock CI (#4705) Geo Min 2026-02-20 11:18:24 -08:00
  • 1f1d0a1ee2 [ci] Adding composablekernel to TheRock CI (#4705) Geo Min 2026-02-20 11:18:24 -08:00
  • 1de8bc9501 Candidate fix 6 Sami Aario 2026-02-20 16:02:27 +00:00
  • 29781f2ac4 [rocm-libraries] ROCm/rocm-libraries#4638 (commit 305ec71) linqunAMD 2026-02-20 15:57:18 +00:00
  • 981c6adfe5 [rocm-libraries] ROCm/rocm-libraries#4638 (commit 305ec71) linqunAMD 2026-02-20 23:56:29 +08:00
  • ad5b05ddcd [ck] Support VGPR estimate in GridwiseGemm_wmma_cshuffle_v3 (#4638) linqunAMD 2026-02-20 23:56:29 +08:00
  • 7a98b5d002 [ck] Support VGPR estimate in GridwiseGemm_wmma_cshuffle_v3 (#4638) linqunAMD 2026-02-20 23:56:29 +08:00
  • a6ffc9c6e5 Candidate fix 5 Sami Aario 2026-02-20 12:30:13 +00:00
  • e2a85ee7a0 Candidate fix 4 Sami Aario 2026-02-20 11:10:25 +00:00
  • 5d40ac6c1b Candidate fix 3 Sami Aario 2026-02-20 10:47:31 +00:00
  • 709608f843 Candidate fix 2 Sami Aario 2026-02-20 10:20:54 +00:00
  • de1a228b34 Candidate fix Sami Aario 2026-02-19 15:44:31 +00:00
  • 8b462b04ce Clear load_tile_transpose_convert_with_offset Sami Aario 2026-02-20 13:07:47 +00:00
  • 1b6bd1986d Adjust test settings for debugging Sami Aario 2026-02-20 13:39:29 +00:00
  • 7689090739 [rocm-libraries] ROCm/rocm-libraries#4556 (commit 15730e7) Aviral Goel 2026-02-20 09:46:22 +00:00
  • 725cd57d43 [rocm-libraries] ROCm/rocm-libraries#4556 (commit 15730e7) Aviral Goel 2026-02-20 13:45:06 +04:00
  • 5d14f3dbce fix: correct ULP calculation in get_absolute_threshold for BF16 tolerance (#4556) Aviral Goel 2026-02-20 13:45:06 +04:00
  • fc7e265563 fix: correct ULP calculation in get_absolute_threshold for BF16 tolerance (#4556) Aviral Goel 2026-02-20 13:45:06 +04:00
  • 99b4da18a2 Update tile_engine/ops/pooling/pool_profiler.hpp ckTileEnginePooling Thomas Ning 2026-02-19 11:18:18 -08:00
  • 9e97c2dc40 Update tile_engine/ops/pooling/pool_instance_builder.py Thomas Ning 2026-02-19 11:17:53 -08:00
  • 7b97e197ef [rocm-libraries] ROCm/rocm-libraries#4299 (commit 668cd49) Márton Bidlek 2026-02-19 08:13:46 +00:00
  • 263288a383 [rocm-libraries] ROCm/rocm-libraries#4299 (commit 668cd49) assistant-librarian[bot] 2026-02-19 09:13:05 +01:00
  • fc19663d91 173 implement device grouped gemm fixed nk for rdna4 (#4299) assistant-librarian[bot] 2026-02-19 09:13:05 +01:00
  • 9c8d3a39ac 173 implement device grouped gemm fixed nk for rdna4 (#4299) assistant-librarian[bot] 2026-02-19 09:13:05 +01:00
  • c5ce5eee5b [rocm-libraries] ROCm/rocm-libraries#4655 (commit f8d76d1) Thrupti Raj Lakshmana Gowda 2026-02-19 06:30:48 +00:00
  • e13d1a228d [rocm-libraries] ROCm/rocm-libraries#4655 (commit f8d76d1) Thrupti Raj Lakshmana Gowda 2026-02-19 00:29:54 -06:00
  • e127f6a9d1 Update CMakeLists.txt (#4655) Thrupti Raj Lakshmana Gowda 2026-02-19 00:29:54 -06:00
  • 33f83b99ed Update CMakeLists.txt (#4655) Thrupti Raj Lakshmana Gowda 2026-02-19 00:29:54 -06:00
  • b2051812bc [rocm-libraries] ROCm/rocm-libraries#4652 (commit 39a5a53) Ville Pietilä 2026-02-19 00:03:19 +00:00
  • 6b9df93342 [rocm-libraries] ROCm/rocm-libraries#4652 (commit 39a5a53) Ville Pietilä 2026-02-19 02:02:13 +02:00
  • 175fce857b Revert "[CK] Add new fwd conv fp16/bf16 instances optimized for unit group size." (#4652) Ville Pietilä 2026-02-19 02:02:13 +02:00
  • 49f62cb57d Revert "[CK] Add new fwd conv fp16/bf16 instances optimized for unit group size." (#4652) Ville Pietilä 2026-02-19 02:02:13 +02:00
  • 0a2b6c4bcd [rocm-libraries] ROCm/rocm-libraries#4297 (commit 5ff580c) Tianxing Wu 2026-02-18 19:33:24 +00:00
  • c49eb518e1 [rocm-libraries] ROCm/rocm-libraries#4297 (commit 5ff580c) assistant-librarian[bot] 2026-02-18 11:32:15 -08:00
  • 72871f5276 moe flatmm xcd remap (#4297) assistant-librarian[bot] 2026-02-18 11:32:15 -08:00
  • 9d437c0630 moe flatmm xcd remap (#4297) assistant-librarian[bot] 2026-02-18 11:32:15 -08:00
  • 5cb8109535 [rocm-libraries] ROCm/rocm-libraries#4640 (commit 37b8c81) Thomas Ning 2026-02-18 15:00:26 +00:00
  • c1517f7590 [rocm-libraries] ROCm/rocm-libraries#4640 (commit 37b8c81) Thomas Ning 2026-02-18 22:59:37 +08:00
  • be25dd6775 Fix the Composable Kernel CI and versions incompatibility (#4640) Thomas Ning 2026-02-18 22:59:37 +08:00
  • e5fb690945 Fix the Composable Kernel CI and versions incompatibility (#4640) Thomas Ning 2026-02-18 22:59:37 +08:00
  • ad36ff239b Add test_load_tile_transpose Sami Aario 2026-01-13 09:04:21 +00:00
  • 1f6768472e [rocm-libraries] ROCm/rocm-libraries#4598 (commit 9ff8af1) John Shumway 2026-02-18 01:27:35 +00:00
  • 7086ee5aea [rocm-libraries] ROCm/rocm-libraries#4598 (commit 9ff8af1) John Shumway 2026-02-17 17:26:32 -08:00
  • 058be6c6e9 [CK_BUILDER] Fix two staging-compiler errors in CK builder code (#4598) John Shumway 2026-02-17 17:26:32 -08:00
  • 24eaadc4d2 [CK_BUILDER] Fix two staging-compiler errors in CK builder code (#4598) John Shumway 2026-02-17 17:26:32 -08:00
  • 2b2a39be98 [rocm-libraries] ROCm/rocm-libraries#4275 (commit 2e07a39) Ville Pietilä 2026-02-18 00:59:15 +00:00
  • 788d66025d [rocm-libraries] ROCm/rocm-libraries#4275 (commit 2e07a39) assistant-librarian[bot] 2026-02-17 17:58:11 -07:00
  • 676ed06e53 [CK] Add new fwd conv fp16/bf16 instances optimized for unit group size. (#4275) assistant-librarian[bot] 2026-02-17 17:58:11 -07:00
  • f0e3f93f1d [CK] Add new fwd conv fp16/bf16 instances optimized for unit group size. (#4275) assistant-librarian[bot] 2026-02-17 17:58:11 -07:00
  • 270b1445b1 [rocm-libraries] ROCm/rocm-libraries#4259 (commit 223d90c) John Shumway 2026-02-17 21:14:11 +00:00
  • 4dc06b5768 [rocm-libraries] ROCm/rocm-libraries#4259 (commit 223d90c) assistant-librarian[bot] 2026-02-17 13:13:19 -08:00
  • 96b64858aa Add multi-file trace parsing and analysis pipeline (#4259) assistant-librarian[bot] 2026-02-17 13:13:19 -08:00
  • a2d139ee59 Add multi-file trace parsing and analysis pipeline (#4259) assistant-librarian[bot] 2026-02-17 13:13:19 -08:00
  • 1bf66006c9 [rocm-libraries] ROCm/rocm-libraries#4272 (commit 52def72) Aviral Goel 2026-02-17 20:42:13 +00:00
  • 7df73fafe0 feat: add new optimized tutorial kernels (#4272) assistant-librarian[bot] 2026-02-17 12:41:06 -08:00
  • ce718ec2e1 feat: add new optimized tutorial kernels (#4272) assistant-librarian[bot] 2026-02-17 12:41:06 -08:00
  • 42973fd546 [rocm-libraries] ROCm/rocm-libraries#4593 (commit a4c2a37) John Shumway 2026-02-17 17:32:55 +00:00
  • 49808ab7b8 [CK_BUILDER] Move some smoke tests that require GPU (#4593) John Shumway 2026-02-17 09:32:15 -08:00
  • cef4904306 [CK_BUILDER] Move some smoke tests that require GPU (#4593) John Shumway 2026-02-17 09:32:15 -08:00
  • f3e4c0721f Improved config. vpietila/retina-net-training-perf Ville Pietilä 2026-02-16 10:48:32 -06:00
  • fb782ba133 Remove additional barriers from double buffer implementation. Ville Pietilä 2026-02-16 09:53:43 -06:00
  • 5494425e7b Double buffering baseline. Ville Pietilä 2026-02-16 07:44:45 -06:00
  • 5f0fd19624 WIP: multi-warp dlejeune/mhc_core Damien Lejeune 2026-02-16 09:33:22 +00:00
  • d20bdfd88b added conv traits to bwd data wmma and wmma v3 instances kabraham/builder_bwd_data Kevin Abraham 2026-02-15 10:43:28 +00:00
  • 2ae886dbe5 added device_grouped_conv_bwd_data_multiple_d_wmma_cshuffle_v3 to builder Kevin Abraham 2026-02-13 15:24:58 +00:00
  • 9c2dd2941b [rocm-libraries] ROCm/rocm-libraries#4419 (commit e241f8b) Jan Patrick Lehr 2026-02-12 22:12:57 +00:00
  • b14f34e97c [CK] Work around staging compiler lifetime warning (#4419) Jan Patrick Lehr 2026-02-12 23:11:53 +01:00
  • c44ccc1d2c [CK] Work around staging compiler lifetime warning (#4419) Jan Patrick Lehr 2026-02-12 23:11:53 +01:00
  • eb6e124e43 added bwd data wmma instances to builder Kevin Abraham 2026-02-12 18:51:37 +00:00
  • dae352e8dc [rocm-libraries] ROCm/rocm-libraries#4282 (commit 2050f93) lalala-sh 2026-02-12 17:45:52 +00:00
  • c41544e621 add memsetasync for ck moe splitk (#4282) assistant-librarian[bot] 2026-02-12 09:44:51 -08:00
  • 015bf06008 add memsetasync for ck moe splitk (#4282) assistant-librarian[bot] 2026-02-12 09:44:51 -08:00
  • 0f55bbae61 [rocm-libraries] ROCm/rocm-libraries#4514 (commit 5378ee0) Illia Silin 2026-02-12 17:19:01 +00:00
  • b5d58b8bc5 [CK] add check for THEROCK_SANITIZER in cmake (#4514) Illia Silin 2026-02-12 09:18:04 -08:00
  • 71d80a6209 [CK] add check for THEROCK_SANITIZER in cmake (#4514) Illia Silin 2026-02-12 09:18:04 -08:00
  • 11d1c40655 V5: experiment with multi-warp Damien Lejeune 2026-02-12 14:39:20 +00:00
  • caa4160b57 WIP sinkhorn Matti Eskelinen 2026-02-12 14:13:14 +00:00
  • 0d7a341d27 V5: reintroduce k-loop + adaptive k-tile size Damien Lejeune 2026-02-12 13:54:58 +00:00
  • 5fe7632393 Add V5: split-k Damien Lejeune 2026-02-12 09:24:15 +00:00
  • 47c7c034e9 [rocm-libraries] ROCm/rocm-libraries#4525 (commit 7f34b22) Illia Silin 2026-02-12 04:43:27 +00:00
  • 14d8dc4714 [CK] Fix the launch_tests script. (#4525) Illia Silin 2026-02-11 20:42:43 -08:00
  • 559a35eee7 [CK] Fix the launch_tests script. (#4525) Illia Silin 2026-02-11 20:42:43 -08:00
  • e1e2f7ac2e [rocm-libraries] ROCm/rocm-libraries#4447 (commit 6d08a99) Christopher Millette 2026-02-11 22:13:15 +00:00
  • d0b17aab8b [CK] Optimize multi-dimensional static for loop decomposition (#4447) Christopher Millette 2026-02-11 15:12:31 -07:00
  • a95cf55aa7 [CK] Optimize multi-dimensional static for loop decomposition (#4447) Christopher Millette 2026-02-11 15:12:31 -07:00
  • ea4942cd02 [rocm-libraries] ROCm/rocm-libraries#4506 (commit d9ccef7) Bartłomiej Kocot 2026-02-11 21:37:50 +00:00
  • 7e1ef9762a Revert "[CK Conv] Add bwd weight instance for large-k shape" (#4506) Bartłomiej Kocot 2026-02-11 22:36:53 +01:00
  • 02afe54a30 Revert "[CK Conv] Add bwd weight instance for large-k shape" (#4506) Bartłomiej Kocot 2026-02-11 22:36:53 +01:00
  • 04eddbc5ce [rocm-libraries] ROCm/rocm-libraries#4471 (commit 10fa702) Christopher Millette 2026-02-11 19:01:05 +00:00
  • 6ed0dde669 [CK] Optimize vector type build times (#4471) Christopher Millette 2026-02-11 11:59:43 -07:00
  • 77169ae227 [CK] Optimize vector type build times (#4471) Christopher Millette 2026-02-11 11:59:43 -07:00
  • 57b036747a WIP: refactor normalisation Damien Lejeune 2026-02-11 16:25:25 +00:00
  • b4112526ce WIP: Triple buffer pipeline. Ville Pietilä 2026-02-11 10:32:35 -05:00
  • 2dd2f114b3 [rocm-libraries] ROCm/rocm-libraries#4407 (commit adde219) Bartłomiej Kocot 2026-02-11 13:43:01 +00:00
  • 4bf06885af [CK][CK TILE] Add has hot loop check for pipeline v1 (#4407) Bartłomiej Kocot 2026-02-11 14:41:59 +01:00
  • 790a786035 [CK][CK TILE] Add has hot loop check for pipeline v1 (#4407) Bartłomiej Kocot 2026-02-11 14:41:59 +01:00
  • 998ddc5d12 Add scheduling barriers and remove debug sync statements. Ville Pietilä 2026-02-11 08:31:56 -05:00
  • 5bacca7b8a Working double buffer implementation for V1 gridwise GEMM pipeline, Ville Pietilä 2026-02-11 06:35:49 -05:00
  • 055de18707 V4: update grid shape to 1D (B) instead of 2D (B,n) Damien Lejeune 2026-02-11 09:46:45 +00:00
  • e88f139c6c [rocm-libraries] ROCm/rocm-libraries#4271 (commit 6fce58e) Johannes Graner 2026-02-11 09:08:38 +00:00