InferWeave

GPUs

Privacy-safe GPU views: card models, fleet counts and per-card availability, with no node identity exposed.

Fixture data (no live surface yet).

Fleet totals

  • GPU cards: 6
  • Available: 3

Card models

  • amd mi300: 2/3 available — available
  • nvidia a100: 0/1 available — fully in use
  • nvidia h100: 1/2 available — available