search
CUDA
Trends
- 1Janus lets a single Go binary run GGUF models on any GPU●Show HN: Janus – Go binary that runs GGUF models via Vulkan on AMD/Intel/Nvidia
A new open-source project called Janus is drawing attention on Hacker News. It is a Go binary that runs GGUF large language models through Vulkan graphics drivers, meaning it works on AMD, Intel and Nvidia GPUs without vendor-specific tooling. Commenters are discussing how it compares to existing inference tools and whether Vulkan can keep pace with CUDA-based performance.
Repos
- magnitudedev/magnitude Open source inference engine for agents that optimizes itself for your exact hardware. Compiles and tunes its kernels on
- tile-ai/tilelang Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels
- antirez/ds4 DeepSeek 4 Flash and PRO local inference engine for Metal, CUDA and ROCm
- lostmsu/TurboGPT Train a tiny GPT in under a minute (CUDA only)