8000 Commits · aakhundov/triton · GitHub

More Web Proxy on the site http://driver.im/

Commits

Commits on Nov 13, 2024

[AMD] Get rid of flat load/store instructions (#5137)
joviliast
authored

Commits on Nov 12, 2024

[AMD] Use DPP to accelerate 16-bit floats (#5072)
joviliast
authored

Commits on Oct 27, 2024

[AMD] Emit vectorized 16-bit float LLVM atomic ops (#4925)
joviliast
authored

Commits on Sep 18, 2024

[AMD] Add load CV cache modifier (#4746)
joviliast
authored

Commits on Sep 12, 2024

[AMD] Preserve order of load instructions when reordering (#4707)
joviliast
authored

Commits on Sep 6, 2024

Revert "[AMD] Disable block merging to avoid block argument explosion (#4176)" (#4631)
joviliast
authored

Commits on Sep 4, 2024

[AMD] Support more cache modifiers for cdna (#4369)
joviliast
authored

Commits on Aug 27, 2024

[AMD] Support emit indices logic WMMAv2 layout (#4518)
joviliast
authored

Commits on Aug 14, 2024

[AMD] Support WMMAv2 dot instruction generation (#4491)
joviliast
authored

Commits on Aug 8, 2024

[AMD][gfx12] Support Shared to dot WMMAv2 operand conversion (#4467)
joviliast
authored

Commits on Aug 2, 2024

[AMD] Support WMMAv2 in AccelerateAMDMatmulPass (#4452)
joviliast
authored

Commits on Aug 1, 2024

[AMD] Introduce 2nd version of WMMA layout (#4292)
joviliast
authored

Commits on Jun 19, 2024

[AMD][NFC] Refactor AccelerateAMDMatmul operand legalization (#4136)
joviliast
authored

Commits on Jun 13, 2024

[AMD][gfx11] Fix BF16 wmma instr generation (#4135)
joviliast
authored

Commits on May 22, 2024

Commits on May 16, 2024

[AMD][Navi31] Apply WMMA for fp8 (#3776)
joviliast
authored

Commits on Apr 18, 2024

[AUTOTUNER] Fix return value of `use_cuda_graph` when `OutOfResources` is raised (#3691)

joviliast
and
Jokeren
authored

Commits on Apr 9, 2024

[AMD][Navi31] Fix WMMA for integer arguments (#3481)

joviliast
and
zahimoud
authored

Commits on Apr 5, 2024

[AMD][Navi31] Fix test_dot for RDNA3 (#3538)
joviliast
authored

Commits on Apr 4, 2024

[AMD] Fix test_annotations.py (#3570)
joviliast
authored

Commits on Mar 26, 2024

[AMD][Navi31] Enable swizzling for WMMA layout (#3404)
joviliast
authored

Commits on Mar 25, 2024

[AMD][Navi] Fix test_convert1d (#3436)
joviliast
authored

Commits on Mar 22, 2024

[AMD][Navi31] Add WMMA 9E7A tests to EmitIndicesTest (#3405)
joviliast
authored

Commits on Mar 15, 2024

[AMD][Navi31] Support WMMA transformations in AccelerateAMDMatmul pass (#3309)
joviliast
authored

Commits on Mar 13, 2024

Commits on Mar 11, 2024

Commits on Mar 8, 2024

[AMD] Disable PipelinePass for AMD target (#3312)
joviliast
authored

Commits on Mar 7, 2024

[AMD] ElementwiseOpToLLVM: Do not convert types if they are equal (#3091)
joviliast
authored

Commits on Mar 6, 2024

[AMD][Navi31] Convert WMMA dot op to LLVM - rebased (#3292)
joviliast
authored

Commits on Mar 5, 2024

Commits on Feb 22, 2024

[AMD][Navi31] Introduce WMMA layout Attr (#3112)
joviliast
authored

0