flitai-io/gpuworks: Lightweight multi-accelerator MaaS platform and cluster manager

GPUWorks is an ultra-lightweight GPU management and Model-as-a-Service platform that combines heterogeneous accelerators across public and private clouds into a single OpenAI-compatible cluster. The entire control plane downloads as a 40 megabyte static binary, featuring a built-in catalog of model launch specs, memory estimation, and container or host runtimes. It supports running side-by-side nodes from NVIDIA, AMD, Huawei Ascend, T-Head PPU, Hygon DCU, and Apple Silicon. Open article →