kv transfer

Compare a complete phase handoff with a tuned combined pool, then test the queueing reversal and qualified completions. Nominal link speed and fleet size alone cannot decide the split.

Long prefills coincide with decode stalls. Compare a tuned combined pool with separate phases, derive transfer and pool capacity, and test whether the improvement survives connector failures.

Compare tensor, pipeline and expert traffic with per-request KV transfer. Work through declared payload and latency assumptions before choosing a serving network.

Size both serving pools from workload-matched rates, then check memory, shared transfer paths and surviving capacity. Follow admission and buffer ownership through failure and scaling.