Data

Tools

Indexes

Applications

Experiments

Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / rocm/composable_kernel issues and pull requests

#2220 - [Issue]: how to use ck as lib in other project

Issue - State: open - Opened by ZJLi2013 2 months ago

#2219 - [CK_TILE] Multiple-D GEMM example

Pull Request - State: open - Opened by mozga-amd 2 months ago

#2218 - Remove not needed bwd wei merged groups instances

Pull Request - State: open - Opened by bartekxk 2 months ago

#2217 - Add IO module for profiler

Issue - State: open - Opened by smedegaard 2 months ago
Labels: enhancement

#2216 - Adds gemm_universal running example in ckProfiler readme

Pull Request - State: closed - Opened by AviralGoelAMD 2 months ago

#2215 - fixed typos in blockwise_gemm_v2

Pull Request - State: closed - Opened by zjing14 2 months ago - 2 comments

#2214 - Refactor tile_window.hpp, tile_window_linear.hpp into a CK Tile Hierarchy

Pull Request - State: closed - Opened by AviralGoelAMD 2 months ago - 1 comment

#2213 - [CK_TILE] Remove extra if from CMakeLists.txt of gemm tests

Pull Request - State: open - Opened by samremes 2 months ago

#2211 - improvements to timing and profiling

Pull Request - State: open - Opened by lfmeadow 2 months ago

#2210 - [CMake] Disable newly added compiler warning -Wnrvo

Pull Request - State: closed - Opened by jplehr 2 months ago - 2 comments

#2209 - [CK_TILE] For FMHA forward kernels, assign block indices reversely if using mask

Pull Request - State: open - Opened by poyenc 2 months ago - 1 comment

#2208 - add int8 support in ck-tile/03_gemm

Pull Request - State: open - Opened by ZJLi2013 2 months ago - 2 comments

#2207 - Fix 11_add_rmsnorm2d_rdquant

Pull Request - State: closed - Opened by SamiAario-AMD 2 months ago

#2206 - Ck tile tutorial examples

Pull Request - State: open - Opened by ClementLinCF 2 months ago

#2205 - [Issue]: int8 gemm pipeline build error

Issue - State: closed - Opened by ZJLi2013 2 months ago - 5 comments

#2204 - Removes debug print statements from CMakeLists files

Pull Request - State: closed - Opened by AviralGoelAMD 3 months ago

#2203 - Fix example_grouped_gemm_multiple_d_xdl_fp16 on gfx950

Pull Request - State: closed - Opened by jefyang1 3 months ago

#2202 - Use new mfma instructions for FP8 on gfx950

Pull Request - State: closed - Opened by jefyang1 3 months ago

#2201 - Restore oddc instances

Pull Request - State: closed - Opened by bartekxk 3 months ago

#2200 - [CK_tile] Add rotating buffer feature for universal gemm

Pull Request - State: open - Opened by amd-khushbu 3 months ago

#2199 - MX GEMM - Expand MX MFMA Testing to BF8, FP6, and BF6 Data Types

Pull Request - State: closed - Opened by andriy-ca 3 months ago - 1 comment

#2196 - Build and store CK library deb package for all targets daily.

Pull Request - State: closed - Opened by illsilin 3 months ago

#2195 - [CK_TILE] Fix fMHA forward batch_prefill kernel codegen errors

Pull Request - State: closed - Opened by poyenc 3 months ago

#2194 - Narrowing error fix for codegen compilation

Pull Request - State: closed - Opened by arai713 3 months ago - 1 comment

#2193 - [Tile Engine] Add benchmark for tile engine gemm.

Pull Request - State: open - Opened by Yanxing-Shi 3 months ago - 3 comments

#2192 - Update the buffer load/store intrinsic names for clang>=20.

Pull Request - State: closed - Opened by illsilin 3 months ago

#2191 - [CK_TILE] Tile loop persistent gemm kernel

Pull Request - State: open - Opened by samremes 3 months ago

#2190 - fix moe sorting build fail

Pull Request - State: closed - Opened by solinzby1 3 months ago
Labels: CI - Pass

#2189 - Adding validation for tile sizes in Tile Engine

Pull Request - State: closed - Opened by amd-khushbu 3 months ago

#2188 - [CK TILE] Grouped Convolution Forward Kernel

Pull Request - State: open - Opened by bartekxk 3 months ago

#2187 - Extend 64x64 with 4 waves instances for grouped conv bwd wei

Pull Request - State: closed - Opened by bartekxk 3 months ago

#2186 - Disable conv for Filter1x1Stride1Pad0 when K or C is even

Pull Request - State: closed - Opened by mozga-amd 3 months ago - 1 comment

#2185 - add CShuffleM/NXdlPerWavePerShuffle in cshuffle_epilogue

Pull Request - State: open - Opened by joyeamd 3 months ago - 3 comments

#2184 - Disable SMFMA gfx90a

Pull Request - State: closed - Opened by amd-khushbu 3 months ago

#2183 - [CK_TILE] Blockwise GEMM pipeline v5

Pull Request - State: open - Opened by aledudek 3 months ago

#2182 - Disable SMFMA for gfx90a

Pull Request - State: closed - Opened by amd-khushbu 3 months ago

#2181 - Switch to v2 pipeline for grouped conv bwd data

Pull Request - State: closed - Opened by bartekxk 3 months ago - 2 comments

#2180 - Generate ckProfiler package for gfx942 only.

Pull Request - State: closed - Opened by illsilin 3 months ago

#2179 - Add grouped conv fwd bias relu instances

Pull Request - State: closed - Opened by bartekxk 3 months ago

#2178 - Add grouped conv fwd bias relu instances

Pull Request - State: closed - Opened by bartekxk 3 months ago

#2177 - transpose load api development

Pull Request - State: open - Opened by joyeamd 3 months ago - 1 comment

#2176 - enable blockwise gemm v3 pipeline in moe

Pull Request - State: open - Opened by lalala-sh 3 months ago

#2175 - Revert "Disable the SMFMA instruction for gfx90a."

Pull Request - State: closed - Opened by ThomasNing 3 months ago

#2174 - Disable the SMFMA instruction for gfx90a.

Pull Request - State: closed - Opened by amd-khushbu 3 months ago

#2173 - Fix grouped conv bwd data tests on gfx950

Pull Request - State: closed - Opened by jefyang1 3 months ago

#2172 - Support for swizzle and transpose for MFMA_16x16x32_F16/BF16

Pull Request - State: closed - Opened by amd-khushbu 3 months ago

#2170 - [DO NOT MERGE] Add a repro example

Pull Request - State: open - Opened by geyyer 3 months ago

#2169 - Revert "Integrate universal gemm with conv bwd data and add SplitK"

Pull Request - State: closed - Opened by aosewski 3 months ago
Labels: urgency_high

#2168 - Flatmm merge

Pull Request - State: closed - Opened by solinzby1 3 months ago

#2167 - [CK_TILE] fix for default epilogue

Pull Request - State: closed - Opened by jakpiase 3 months ago

#2165 - Move 16x16 grouped conv fwd instances from comp header

Pull Request - State: closed - Opened by bartekxk 3 months ago

#2164 - [LWPCK-2957] fix moe gemm2 for gfx950

Pull Request - State: closed - Opened by mtgu0705 3 months ago

#2162 - Support fp8 in example tile_example_flatmm_basic

Pull Request - State: closed - Opened by linqun 3 months ago

#2159 - [CK_Tile] Simplified Mem pipeline

Pull Request - State: closed - Opened by amd-khushbu 3 months ago - 1 comment

#2157 - Restrict MX GEMM Instantiation to GFX950 Architecture

Pull Request - State: closed - Opened by andriy-ca 3 months ago

#2151 - Add FP4 MX MFMA tests

Pull Request - State: closed - Opened by geyyer 3 months ago

#2146 - [CK_TILE] Grouped GEMM tile loop

Pull Request - State: closed - Opened by samremes 3 months ago - 1 comment

#2144 - [DO NOT MERGE] WIP Bgp v5

Pull Request - State: closed - Opened by aledudek 3 months ago

#2141 - Run CI jobs as user jenkins

Pull Request - State: closed - Opened by illsilin 3 months ago

#2140 - Add grouped conv fwd 16x16 mfma instruction instances

Pull Request - State: closed - Opened by bartekxk 3 months ago

#2139 - fix moe i4 example bug

Pull Request - State: closed - Opened by lalala-sh 3 months ago

#2138 - [CK Tile] Fix the numerical issue for MFMA 16x16x128 for FP8

Pull Request - State: closed - Opened by DDEle 3 months ago

#2137 - enable hd128 swa

Pull Request - State: closed - Opened by slippedJim 3 months ago

#2135 - [Tile Engine] Add benchmark for tile engine

Pull Request - State: closed - Opened by Yanxing-Shi 3 months ago - 3 comments

#2134 - [Tile Engine] Improved README.md

Pull Request - State: closed - Opened by AviralGoelAMD 3 months ago - 1 comment
Labels: documentation

#2131 - Vectorized Transpose for Batched Transpose CK Tile Operator

Pull Request - State: closed - Opened by ThomasNing 3 months ago - 2 comments

#2125 - Support for MFMA_16x16x128 for fp8/bf8

Pull Request - State: closed - Opened by amd-khushbu 3 months ago

#2114 - [CK_TILE] FMHA Support hdim_v to as a Multiple of 32

Pull Request - State: open - Opened by DDEle 3 months ago

#2113 - Update Code Ownership

Pull Request - State: closed - Opened by ThomasNing 3 months ago

#2112 - Temporarily disable MX FP4 device tests

Pull Request - State: closed - Opened by geyyer 3 months ago

#2105 - fix pk_i4_v3 tests failures in Unbuntu env.

Pull Request - State: closed - Opened by mtgu0705 3 months ago - 1 comment

#2104 - fix CI build fail

Pull Request - State: closed - Opened by solinzby1 3 months ago

#2103 - MFMA 16x16x32fp8

Pull Request - State: closed - Opened by ThomasNing 3 months ago

#2102 - GEMM Multiply Multiply Fix

Pull Request - State: closed - Opened by ThomasNing 3 months ago

#2101 - Floating Point Exception Error fix for Streamk GEMM Kernel

Pull Request - State: closed - Opened by ozturkosu 3 months ago

#2098 - fix for basic gemm UseStructuredSparsity

Pull Request - State: closed - Opened by jakpiase 3 months ago - 1 comment

#2097 - [Issue]: Building CK fails due to -fuse-ld=path deprecation

Issue - State: open - Opened by esna-echo 3 months ago - 2 comments
Labels: Under Investigation

#2087 - Added a much simpler example and commented 01_add

Pull Request - State: closed - Opened by AviralGoelAMD 4 months ago - 2 comments

#2081 - [Performance] AMD wmma redundant load issue

Issue - State: open - Opened by DoubleClark 4 months ago

#2080 - [Draft] For multi instance generation for CkTileEngine

Pull Request - State: open - Opened by amd-khushbu 4 months ago

#2078 - addded documentation for ck_tile::array<T,N>

Pull Request - State: open - Opened by AviralGoelAMD 4 months ago
Labels: documentation

#2077 - Fix build issues for multiple targets.

Pull Request - State: closed - Opened by illsilin 4 months ago

#2076 - [Issue]: Building the CK library fails to complete

Issue - State: open - Opened by stefan0re 4 months ago - 1 comment
Labels: Under Investigation

#2075 - CK pk_i4_t test failures fix

Pull Request - State: closed - Opened by mtgu0705 4 months ago