cubecl

Minus the NPU target, if you want Mojo in Rust there is https://github.com/tracel-ai/cubecl

For compute shaders, you can already do this with CubeCL: https://github.com/tracel-ai/cubecl You write kernels in a Rust DSL using #[cube], it supports WebGPU through WGSL and Vulkan through SPIR-V, along with CUDA, AMD via ROCm, and Metal. (disclosure: I am a contributor)

Burn's great, and we cite it in the paper's related work. Different layer, though: Burn is the high-level DL framework (tensor ops, autodiff, backends), whereas cuTile Rust is the kernel-authoring layer underneath. The closer neighbor is actually CubeCL, Tracel's compute layer that Burn sits on, and there's an open issue to integrate cuTile Rust into CubeCL, so we're already looking at how these fit together. See https://github.com/tracel-ai/cubecl/issues/1282

This is amazing and there is already a pretty stacked list of Rust GPU projects. This seems to be at an even lower level of abstraction than burn[0] which is lower than candle[1]. I gueds whats left is to add backend(s) that leverage naga and others to the above projects? Feeks like everyone is building on different bases here, though I know the naga work is relatively new. [EDIT] Just to note, burn is the one that focuses most on platform support but it looks like the only backend that uses naga is wgpu... So just use wgpu and it's fine? Yeah basically wgpu/ash (vulkan, metal) or cuda [EDIT2] Another crate closer to this effort: https://github.com/tracel-ai/cubecl [0]: https://github.com/tracel-ai/burn [1]: https://github.com/huggingface/candle/

Agreed! I was looking through the summation example < https://github.com/tracel-ai/cubecl/blob/main/examples/sum_t...> and it seems like the primary focus is on the more traditional pre-2018 GPU programming without explicit warp-level operations, asynchrony, atomics, barriers, or countless tensor-core operations. The project feels very nice and it would be great to have more notes in the README on the excluded functionality to better scope its applicability in more advanced GPGPU scenarios.

In terms of speed, Rust is up there with C/C++. See e.g. https://benchmarksgame-team.pages.debian.net/benchmarksgame/... I also ported several algo from C and it was matching the performance. Regarding SIMD support, the only thing that is missing, is stable support for avx512, and some more exotic feature extensions for deep learning e.g. avx_vnni. Those are implemented and waiting to be included in the next stable versions. Gpu support: this is still an issue b/c of not enough people working on it, but there projects trying to improve this: see https://github.com/tracel-ai/cubecl . Const generics: Yeah, there are a few annoying issues: it is limited to small set of types. For instance, you cant use const enum as a generic. Also, you cant use generic parameters in const operations on stable rust: see unstable feature generic_const_exprs.

Available (and planned) runtimes: https://github.com/tracel-ai/cubecl?tab=readme-ov-file#runti...