model
Model used in the legacy InferenceX static 8K/1K sequence length scenario where GB300 NVL72 leads H200 by up to 40x tokens per megawatt.
No current published clusters are mapped here yet.