Ecosyste.ms: Issues

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

GitHub / JuliaGPU/GemmKernels.jl issues and pull requests

#198 - CompatHelper: bump compat for LLVM to 8, (keep existing compat)

Pull Request - State: closed - Opened by github-actions[bot] 3 months ago - 1 comment

#196 - Add new pipelined kernel

Pull Request - State: open - Opened by thomasfaingnaert 5 months ago

#195 - Add support for CTA swizzling

Pull Request - State: closed - Opened by thomasfaingnaert 5 months ago

#194 - Use a zero layout for C in shared memory if beta=0

Pull Request - State: closed - Opened by thomasfaingnaert 5 months ago

#193 - Fix vstorea! not being inlined

Pull Request - State: closed - Opened by thomasfaingnaert 5 months ago

#192 - Consider a patch release for the new CUDA versions?

Issue - State: closed - Opened by avik-pal 6 months ago - 7 comments

#191 - Bump julia-actions/setup-julia from 1 to 2

Pull Request - State: closed - Opened by dependabot[bot] 6 months ago - 2 comments
Labels: dependencies

#190 - Refactor tuning script

Pull Request - State: closed - Opened by maleadt 8 months ago - 2 comments

#189 - Resolve remaining issues with benchmarking

Issue - State: open - Opened by thomasfaingnaert 8 months ago

#188 - Skip configurations with fewer than 4 warps in tuning

Pull Request - State: open - Opened by thomasfaingnaert 8 months ago - 2 comments

#187 - Remove Julia 1.8 from CI

Pull Request - State: closed - Opened by thomasfaingnaert 8 months ago

#186 - Get benchmarks working again

Pull Request - State: closed - Opened by thomasfaingnaert 8 months ago - 9 comments

#185 - Apply isapprox elementwise

Pull Request - State: closed - Opened by thomasfaingnaert 8 months ago - 1 comment

#183 - Extend set of WMMA operator shapes

Pull Request - State: closed - Opened by thomasfaingnaert 10 months ago - 1 comment

#182 - FPUOp: Ensure the FMA operator is inlined.

Pull Request - State: closed - Opened by maleadt 10 months ago - 2 comments

#181 - Check size limits of LocalArray

Pull Request - State: open - Opened by thomasfaingnaert 10 months ago - 2 comments

#180 - Check tile sizes in config

Pull Request - State: closed - Opened by thomasfaingnaert 10 months ago - 2 comments

#179 - Add script to tune parameters

Pull Request - State: closed - Opened by thomasfaingnaert 10 months ago - 2 comments

#178 - Fix typo in parallelise function name

Pull Request - State: closed - Opened by thomasfaingnaert 11 months ago - 2 comments

#177 - A wrong function name `parallellise`

Issue - State: closed - Opened by ArrogantGao 11 months ago - 1 comment

#176 - Do not hardcode vectorisation width in layouts

Pull Request - State: closed - Opened by thomasfaingnaert 11 months ago - 2 comments

#175 - Fix alignment check for non 16-byte alignments

Pull Request - State: closed - Opened by thomasfaingnaert 11 months ago - 2 comments

#174 - Check number of threads before launching kernel

Pull Request - State: closed - Opened by thomasfaingnaert 11 months ago - 3 comments

#173 - Check number of stages for pipelined kernel

Pull Request - State: closed - Opened by thomasfaingnaert 11 months ago - 3 comments

#172 - Improve heuristic for memcopy tile sizes

Pull Request - State: closed - Opened by thomasfaingnaert 11 months ago - 2 comments

#171 - Test more WMMA configurations

Pull Request - State: closed - Opened by thomasfaingnaert 11 months ago - 1 comment

#170 - Refactor configs to use macros

Pull Request - State: closed - Opened by thomasfaingnaert 11 months ago - 1 comment

#169 - Compare with cuBLAS during benchmarking

Pull Request - State: closed - Opened by thomasfaingnaert 11 months ago - 2 comments

#168 - Adapt to CUDA.jl profile changes

Pull Request - State: closed - Opened by thomasfaingnaert 11 months ago - 1 comment

#167 - Add a check for the block shape in the K dimension

Pull Request - State: closed - Opened by wardvermeulen 11 months ago

#166 - Throw ConfigError for unsupported WMMA shapes

Pull Request - State: closed - Opened by thomasfaingnaert 11 months ago - 2 comments

#165 - FPU operator issues

Issue - State: open - Opened by maleadt 11 months ago - 3 comments

#164 - Fix configuration heuristic.

Pull Request - State: closed - Opened by maleadt 11 months ago - 2 comments

#163 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] 12 months ago - 1 comment

#162 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] 12 months ago - 2 comments

#161 - Add more flexible FPU operator

Pull Request - State: closed - Opened by wardvermeulen 12 months ago - 1 comment

#160 - Rework benchmarks and tests

Pull Request - State: closed - Opened by thomasfaingnaert 12 months ago - 2 comments

#159 - Adding more Semirings

Issue - State: open - Opened by Wimmerer 12 months ago - 1 comment

#158 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] about 1 year ago - 2 comments

#157 - Bump actions/checkout from 3 to 4

Pull Request - State: closed - Opened by dependabot[bot] about 1 year ago - 2 comments
Labels: dependencies

#156 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] about 1 year ago - 2 comments

#155 - CompatHelper: bump compat for "CUDA" to "5"

Pull Request - State: closed - Opened by github-actions[bot] about 1 year ago - 4 comments

#154 - Parameter tuning

Issue - State: open - Opened by maleadt about 1 year ago - 1 comment

#153 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] about 1 year ago - 2 comments

#152 - Questions about usage of registers

Issue - State: closed - Opened by ArrogantGao about 1 year ago - 3 comments

#151 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] about 1 year ago - 2 comments

#150 - Bump actions/checkout from 2 to 3

Pull Request - State: closed - Opened by dependabot[bot] about 1 year ago - 1 comment
Labels: dependencies

#149 - Bump peter-evans/create-pull-request from 3 to 5

Pull Request - State: closed - Opened by dependabot[bot] about 1 year ago - 1 comment
Labels: dependencies

#148 - enable dependabot for GitHub actions

Pull Request - State: closed - Opened by ranocha about 1 year ago - 3 comments

#147 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] about 1 year ago - 2 comments

#146 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] about 1 year ago - 2 comments

#145 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] about 1 year ago - 2 comments

#144 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] about 1 year ago - 2 comments

#143 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] about 1 year ago - 2 comments

#142 - Replace LocalArray with SArray

Issue - State: open - Opened by maleadt about 1 year ago

#141 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] about 1 year ago - 1 comment

#140 - Use cached loads

Issue - State: open - Opened by maleadt about 1 year ago

#139 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] about 1 year ago - 1 comment

#138 - Show kernel details on benchmark differences.

Pull Request - State: closed - Opened by maleadt about 1 year ago - 1 comment

#137 - Add a mechanism to expose execution details to callers.

Pull Request - State: closed - Opened by maleadt about 1 year ago - 2 comments

#136 - Simplify config definition and usage.

Pull Request - State: closed - Opened by maleadt about 1 year ago - 3 comments

#135 - Check if the warp doesn't index out of the tile subpartition.

Pull Request - State: closed - Opened by maleadt about 1 year ago - 2 comments

#134 - Detect alignment issues and throw a Julia error.

Pull Request - State: closed - Opened by maleadt about 1 year ago - 2 comments

#133 - Add example.

Pull Request - State: closed - Opened by maleadt about 1 year ago - 2 comments

#132 - Put the BLAS interface directly in the GemmKernels.jl module.

Pull Request - State: closed - Opened by maleadt about 1 year ago - 2 comments

#131 - Fix fragtypes of ColMajor and RowMajor fallback layouts.

Pull Request - State: closed - Opened by maleadt about 1 year ago - 2 comments

#130 - Add layouts for accessing unaligned or non tile-sized global.

Pull Request - State: closed - Opened by maleadt about 1 year ago - 3 comments

#129 - BLAS: Convert alpha & beta to more appropriate types.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 2 comments

#128 - Restrict scope of VecElement usage.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 2 comments

#127 - Fix vector op indexing and add boundscheck.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 4 comments

#126 - Update manifest

Pull Request - State: closed - Opened by github-actions[bot] over 1 year ago - 2 comments

#125 - Use Octavian.jl for large mixed-mode CPU calculations.

Pull Request - State: open - Opened by maleadt over 1 year ago - 7 comments

#124 - Simplify tests.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 3 comments

#123 - Transform VecElement-contained values.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 2 comments

#122 - Unify WMMA and FPU operator typevars [NFC]

Pull Request - State: closed - Opened by maleadt over 1 year ago - 2 comments

#121 - Use XUnit.jl for parallel testing.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 3 comments

#120 - Add zero layout to optimize alpha/beta=zero.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 2 comments

#119 - Introduce a helper macro to simplify immutable indexing.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 2 comments

#118 - Commit the Manifest.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 2 comments

#117 - Improve benchmarks

Issue - State: open - Opened by maleadt over 1 year ago

#116 - Add a benchmarks bot.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 2 comments

#115 - Benchmark bot

Issue - State: closed - Opened by maleadt over 1 year ago - 2 comments

#114 - Transform functions: pass values, not VecElements

Issue - State: closed - Opened by maleadt over 1 year ago

#113 - Enable use of FPU operator in BLAS wrappers.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 4 comments

#112 - Configure and check shared memory automatically.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 2 comments

#111 - Make vectorized store convert and perform multiple stores if required

Pull Request - State: closed - Opened by maleadt over 1 year ago - 3 comments

#110 - Optimizations when alpha or beta is 0

Issue - State: closed - Opened by maleadt over 1 year ago - 2 comments

#109 - Make LocalArray setindex convert.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 1 comment

#108 - Using GemmKernels.jl in CUDA.jl

Issue - State: open - Opened by maleadt over 1 year ago

#107 - Replace KernelAbstractions with LLVMLoopInfo.

Pull Request - State: closed - Opened by maleadt over 1 year ago - 2 comments

#106 - CompatHelper: bump compat for LLVM to 6, (keep existing compat)

Pull Request - State: closed - Opened by github-actions[bot] over 1 year ago - 1 comment

#105 - Tensor contractions

Pull Request - State: open - Opened by wardvermeulen over 1 year ago - 4 comments

#104 - Use LLVMLoopInfo.jl

Issue - State: closed - Opened by maleadt over 1 year ago - 1 comment

#103 - Bump compat bounds to use newer CUDA.jl

Pull Request - State: closed - Opened by maleadt over 1 year ago - 1 comment

#102 - Add CI for Julia 1.9

Pull Request - State: closed - Opened by thomasfaingnaert over 1 year ago - 1 comment

#101 - FPU operator

Pull Request - State: closed - Opened by wardvermeulen over 1 year ago - 4 comments

#99 - Large LocalArray eltypes runs into compiler heuristics

Issue - State: open - Opened by maleadt about 2 years ago - 5 comments

#98 - Re-land StaticArrays removal

Pull Request - State: closed - Opened by maleadt about 2 years ago - 1 comment