Nvidia brings native GPU programming to Rust with CUDA Rust
Nvidia announces native GPU programming in Rust
Nvidia has announced CUDA Rust, enabling developers to write GPU kernels directly in Rust and compile them natively to PTX. The toolchain offers two tracks: SIMT, for thread-level programming like CUDA C++, and Tile, a higher-level model where the compiler maps tiles to hardware. Both aim to catch memory and aliasing bugs at compile time, with inter-language interop planned. The move reflects Rust's growing role in AI systems infrastructure.
The GPU kernel is the exception. You can launch kernels from Rust, but the kernel itself often has to be written in another language. NVIDIA CUDA Rust closes that gap.