Topic
Splitting AI workloads across multiple model and accelerator types rather than a uniform architecture.
No current published clusters are mapped here yet.