If you run vLLM with --kv-cache-dtype fp8 on a DeepSeek-family (MLA) model and your GPU is a GB10, an RTX PRO 6000, or any workstation or consumer Blackwell card, there is a decent chance the engine dies on startup with:

triton.runtime.errors.OutOfResources: out of resource: shared memory,

Required: 102400, Hardware limit: 101376. Reducing block sizes or `num_stages` may help.

Enter fullscreen mode

Exit fullscreen mode