Show HN: Phobos – A tiny scale-free kernel language with tile-DAG support
4 days ago
- #GPU Programming
- #Machine Learning
- #Compiler Design
- The author developed Phobos, a tiny kernel language inspired by Triton that compiles to PTX and runs on NVIDIA GPUs.
- Phobos achieves about 76% of cuBLAS SGEMM performance on an RTX 2080 SUPER, with optimizations like tiling and tensor core usage.
- The language abstracts thread-level details, focusing on tile-level operations for AI computations, enabling natural scaling to distributed clusters.
- The project involved learning low-level GPU programming, using MLIR and LLVM with Rust, and implementing autotuning for kernel parameters.
- A cluster prototype was built for distributed tile-DAG execution, though multi-node benchmarks were not performed due to hardware limitations.