The Wayback Machine - https://web.archive.org/web/20230615003407/https://github.com/antiagainst

antiagainst

Follow

Lei Zhang antiagainst

Follow

AI Frameworks & Compilers. Now: Vulkan compute, IREE, MLIR. Previous: Vulkan graphics, SPIR-V toolchain.

880 followers · 11 following

Achievements

BetaSend feedback

Achievements

BetaSend feedback

Organizations

Block or Report

Block or report antiagainst

Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse

Pinned

gfx-rs/rspirv Public

Rust implementation of SPIR-V module processing functionalities

Rust 402 62
google/uVkCompute Public

A micro Vulkan compute pipeline and a collection of benchmarking compute shaders

C++ 145 28
openxla/iree Public

A retargetable MLIR-based machine learning compiler and runtime toolkit.

C++ 1.9k 431
llvm/llvm-project Public

The LLVM Project is a collection of modular and reusable compiler and toolchain technologies. Note: the repository does not accept github pull requests at this moment. Please submit your patches at…

20.2k 7.5k
microsoft/DirectXShaderCompiler Public

This repo hosts the source for the DirectX Shader Compiler which is based on LLVM/Clang.

C++ 2.6k 606
KhronosGroup/SPIRV-Tools Public

C++ 881 482

1,483 contributions in the last year

Learn how we count contributions

Activity overview

Contributed to openxla/iree, llvm/llvm-project, Kapeli/Dash-User-Contributions and 16 other repositories

Contribution activity

June 2023

Created 37 commits in 1 repository

openxla/iree 37 commits

Created a pull request in openxla/iree that received 29 comments

[cuda] Implement basics for a CUDA HAL driver rewrite

This commit starts a CUDA HAL driver rewrite under experimental/. We create a new cuda2/ directory to host the new code to avoid interrupting the c…

+1,556 −0 • 29 comments

Opened 20 other pull requests in 2 repositories

openxla/iree 17 merged 1 closed 1 open

Kapeli/Dash-User-Contributions 1 open

Update Vulkan docset to version 1.3.253 Jun 13

Reviewed 34 pull requests in 1 repository

openxla/iree 25 pull requests

[StableHLO] Use stablehlo submodule Jun 14
[metal] Implement a Metal HAL driver Jun 14
Use cuGetProcAddress to load CUDA entry points Jun 13
Fuse iota ops with consumers always. Jun 12
Reword comments for IREE_BUILD_DOCS. Jun 12
Allow defining IREE_HOST_SIZE_T to other types. Jun 12
[spirv] Dump spirv.module with dump-executable-intermediates-to= Jun 12
[StableHLO] Make reduce lowering more robust Jun 12
Correctly tag Vulkan Ampere tests as requiring sm80 Jun 12
[GPUCheckResourceUsage] Don't choke on alloc of memref of index Jun 12
Cleaning up the tracing.h mechanism to enable alternative implementations. Jun 10
[cuda] Port over descriptor set and pipeline layout Jun 9
Allowing for sync allocations to be deallocated via queue-ordered deallocas. Jun 9
Adding a fallback for when CUDA memory pools are unsupported. Jun 9
[ConvertToLLVM] Don't choke on alloc of memref of index Jun 9
Refresh deployment-configuration website pages. Jun 8
Adding IREE_HAL_EXTERNAL_BUFFER_TYPE_DEVICE_ALLOCATION. Jun 8
Remove MHLO support Jun 8
Retain the parent channel on a split in iree_hal_nccl_channel_t. Jun 8
Moving builtins lower in the pipeline and adding option to force. Jun 8
Adding support for async memory pool allocations in the CUDA HAL. Jun 8
[cuda] Port over allocator and buffer implementation Jun 8
Integrate llvm-project at 223a0f63 Jun 7
[ci] Update NVIDIA driver packages to v530 in docker images Jun 5
[cuda] Implement basics for a CUDA HAL driver rewrite Jun 5
Some pull request reviews not shown.

Created an issue in openxla/iree that received 1 comment

Use past average latency for comparison on pull request benchmarks

Right now we are using the last landed commit's latency for comparison when performing benchmarks on pull requests. With just one single data point…

1 comment

Opened 4 other issues in 1 repository

openxla/iree 4 open