other
Groq's custom inference chip architecture, licensed non-exclusively to Nvidia.
No evidence-backed relationships are recorded.
No current published clusters are mapped here yet.