https://github.com/termoshtt/accel
(Mirror of GitLab) GPGPU Framework for Rust
https://github.com/termoshtt/accel
cuda gpgpu rust-lang
Last synced: 3 months ago
JSON representation
(Mirror of GitLab) GPGPU Framework for Rust
- Host: GitHub
- URL: https://github.com/termoshtt/accel
- Owner: termoshtt
- License: other
- Archived: true
- Created: 2017-11-01T11:46:43.000Z (over 7 years ago)
- Default Branch: master
- Last Pushed: 2020-09-30T16:29:57.000Z (over 4 years ago)
- Last Synced: 2024-05-02T00:36:56.719Z (12 months ago)
- Topics: cuda, gpgpu, rust-lang
- Language: Rust
- Homepage: https://gitlab.com/termoshtt/accel
- Size: 1.09 MB
- Stars: 443
- Watchers: 27
- Forks: 17
- Open Issues: 10
-
Metadata Files:
- Readme: README.md
- Changelog: CHANGELOG.md
- Funding: .github/FUNDING.yml
- License: LICENSE-APACHE
Awesome Lists containing this project
README
Accel: GPGPU Framework for Rust
================================
[](https://gitlab.com/termoshtt/accel/commits/master)
|crate |crates.io |docs.rs |GitLab Pages | |
|:-----------|:---------------------------------------------------------------------------|:----------------------------------------------------------------------|:-----------------------------------------------------------------------------|:------------------------------------------|
|accel |[][crate/accel] |[][docs/accel] |[][dev/accel] |CUDA-based GPGPU framework |
|accel-core |[][crate/accel-core] |[][docs/accel-core] |[][dev/accel-core] |Helper for writing device code |
|accel-derive|[][crate/accel-derive]|[][docs/accel-derive]|[][dev/accel-derive]|Procedural macro for generating kernel code|
[crate/accel]: https://crates.io/crates/accel/0.3.0
[crate/accel-core]: https://crates.io/crates/accel-core/0.3.0
[crate/accel-derive]: https://crates.io/crates/accel-derive/0.3.0
[docs/accel]: https://docs.rs/accel/0.3.0
[docs/accel-core]: https://docs.rs/accel-core/0.3.0
[docs/accel-derive]: https://docs.rs/accel-derive/0.3.0
[dev/accel]: https://termoshtt.gitlab.io/accel/accel/accel
[dev/accel-core]: https://termoshtt.gitlab.io/accel/accel/accel_core
[dev/accel-derive]: https://termoshtt.gitlab.io/accel/accel/accel_derive
Requirements
------------
- Minimum Supported Rust Version (MSRV) is 1.42.0
- Install [CUDA](https://developer.nvidia.com/cuda-downloads) on your system
- accel depends on CUDA Device APIs through [rust-cuda/cuda-sys](https://github.com/rust-cuda/cuda-sys)
- accel does not depend on CUDA Runtime APIs. It means that a compiled binary requires only `libcuda.so` at runtime, which is far lighter than entire CUDA development toolkit.
- Setup NVPTX target of Rust
- Install `nightly-2020-05-01` toolchain with `nvptx64-nvidia-cuda` target, and [rust-ptx-linker](https://github.com/denzp/rust-ptx-linker)
- There is an [setup script](setup_nvptx_toolchain.sh) for them:
```
curl -sSL https://gitlab.com/termoshtt/accel/raw/master/setup_nvptx_toolchain.sh | bash
```
Or, you can use [docker container](./docker)
Limitations
------------
This project is still in early stage. There are several limitations as following:
- For runtime on CPU
- [Windows](https://gitlab.com/termoshtt/accel/-/issues/25) and macOS are not supported
- [f64](https://gitlab.com/termoshtt/accel/-/issues/53) and [Complex number](https://gitlab.com/termoshtt/accel/-/issues/54) supports are missing
- [Texture/Surface object handling](https://gitlab.com/termoshtt/accel/-/issues/40) is missing
- Async features based on CUDA Stream and Events are disabled until [async/.await support](https://gitlab.com/termoshtt/accel/-/issues/4)
- For writing GPU kernel code
- [libstd cannot be used in writing kernel](https://gitlab.com/termoshtt/accel/-/issues/38)
- [Rust slice cannot be used in writing kernel](https://gitlab.com/termoshtt/accel/-/issues/7)
- [Shared memory](https://gitlab.com/termoshtt/accel/-/issues/39) cannot be used
Contribution
------------
This project is developed on [GitLab](https://gitlab.com/termoshtt/accel) and mirrored to [GitHub](https://github.com/rust-accel/accel).
Sponsors
--------
- [RICOS Co. Ltd](https://www.ricos.co.jp/)
- GPU instances for CI and development
Links
------
Projects which accel depends on:
- [rust-cuda/cuda-sys](https://github.com/rust-cuda/cuda-sys): CUDA Runtime and Driver API binding to Rust
- [denzp/rust-ptx-linker](https://github.com/denzp/rust-ptx-linker): Linker for PTX files generated by `rustc`
Related Projects:
- [rust-cuda/wg](https://github.com/rust-cuda/wg): Working group for Rust CUDA Team
- [denzp/rust-ptx-builder](https://github.com/denzp/rust-ptx-builder): Another CUDA kernel builder from Rust crate
- [bheisler/RustaCUDA](https://github.com/bheisler/RustaCUDA): Another CUDA-based Rust framework