CubeCL: GPU Kernels in Rust for CUDA, ROCm, and WGPU
1–10 of 45 posts
Re: CubeCL: GPU Kernels in Rust for CUDA, ROCm, and WGPU
#2Very interesting project! I am wondering how it compare against OpenCL, which I think adopts the same fundamental idea (write once, run everywhere)? Is it about CUbeCL's internal optimization for Rust that happens at compile time?
Re: CubeCL: GPU Kernels in Rust for CUDA, ROCm, and WGPU
#3See also this overview for how it compares to other projects in the Rust and GPU ecosystem: https://rust-gpu.github.io/ecosystem/
Re: CubeCL: GPU Kernels in Rust for CUDA, ROCm, and WGPU
#4See also this overview for how it compares to other projects in the Rust and GPU ecosystem: https://rust-gpu.github.io/ecosystem/
Surprised this doesn't mention candle: https://github.com/huggingface/candle
Re: CubeCL: GPU Kernels in Rust for CUDA, ROCm, and WGPU
#5Very interesting project! I am wondering how it compare against OpenCL, which I think adopts the same fundamental idea (write once, run everywhere)? Is it about CUbeCL's internal optimization for Rust that happens at compile time?
A lot of things happen at compile time, but you can execute arbitrary code in your kernel that executes at compile time, similar to generics, but with more flexibility. It's very natural to branch on a comptime config to select an algorithm.
Re: CubeCL: GPU Kernels in Rust for CUDA, ROCm, and WGPU
#6Love it. I've been using cudarc lately; would love to try this since it looks like it can share data structures between host and device (?). I infer that this is a higher-level abstraction.
Re: CubeCL: GPU Kernels in Rust for CUDA, ROCm, and WGPU
#7See also this overview for how it compares to other projects in the Rust and GPU ecosystem: https://rust-gpu.github.io/ecosystem/
Surprised this doesn't mention candle: https://github.com/huggingface/candle
I don't think that fits; that's a ML framework. The others in the link are general GPU frameworks.
Re: CubeCL: GPU Kernels in Rust for CUDA, ROCm, and WGPU
#8Very interesting project! I am wondering how it compare against OpenCL, which I think adopts the same fundamental idea (write once, run everywhere)? Is it about CUbeCL's internal optimization for Rust that happens at compile time?
A lot of things happen at compile time, but you can execute arbitrary code in your kernel that executes at compile time, similar to generics, but with more flexibility. It's very natural to branch on a comptime config to select an algorithm.
[dead]
Re: CubeCL: GPU Kernels in Rust for CUDA, ROCm, and WGPU
#9Where is the Metal love…
Re: CubeCL: GPU Kernels in Rust for CUDA, ROCm, and WGPU
#10Where is the Metal love…
It also compiles directly to MSL, it is just missing from the post title.