← Research
Scheduling · Planned
Scheduling Across Heterogeneous QPUs
How should infrastructure place and schedule stages when backends differ in availability, topology, queue time, fidelity, cost and policy?
Heterogeneous QPUs break naive portability. Placement research must separate local intent from remote provider queues and avoid pretending backends are interchangeable.
We are researching decision inputs — including when a simulator is the better target than a QPU — not shipping an automatic multi-QPU scheduler today.
Status: planned.