Burn is a next generation tensor library and Deep Learning Framework that doesn't compromise on flexibility, efficiency and portability.

CUDA on non-NVIDIA GPUs

Ecosystem of libraries and tools for writing and executing fast GPU code fully in Rust.

cuda-oxide is an experimental Rust-to-CUDA compiler that lets you write (SIMT) GPU kernels in safe(ish), idiomatic Rust. It compiles standard Rust code directly to PTX — no DSLs, no foreign language bindings, just Rust.

Apache Mahout - an environment for quickly creating scalable, performant machine learning applications.

Multi-platform high-performance compute language extension for Rust.

Deep learning in Rust, with shape checked tensors and neural networks

Safe rust wrapper around CUDA toolkit

The Hacker's Machine Learning Engine

A machine learning library with a TypeScript API and Rust backend. CUDA and WebGPU compatibility. Built to understand how ML frameworks and models work internally.

Rust wrapper for ArrayFire

Rusty wrapper for the CUDA Driver API

cuTile Rust provides a safe, tile-based kernel programming DSL for the Rust programming language. It features a safe host-side API for passing tensors to asynchronously executed kernel functions.

Pure Rust Inference Engine

Pure Rust + CUDA LLM inference engine — no PyTorch, OpenAI-compatible, serves Qwen3 to Kimi-K2

Demo project showing a single Rust codebase running on CPU and directly on GPUs

(Mirror of GitLab) GPGPU Framework for Rust

A Rust library integrated with ONNXRuntime, providing a collection of Computer Vison and Vision-Language models such as YOLO, FastVLM, and more.

A machine learning library for Rust.

From-scratch Rust+CUDA inference engine, bit-exact by construction — NVFP4, MoE, MTP speculative decoding, tuned against measured limits of one RTX 5090 Laptop (sm_120a).