Determine the complexity of minimum-action seating

Determine the computational complexity of energy-optimal layout, or minimum-action seating, for programs on cyclic-storage machines, including the additional scheduling dimension absent from bandwidth minimization; in particular, resolve the conjecture that the problem is NP-hard.

Background

The cyclic-storage model represents programs by packets placed on rotating rings, with energy determined by the action of live packets, transfers, and parked storage. Earlier results establish constant-factor inapproximability for a related minimum-window seating problem, but they do not settle the complexity of minimizing the full energy/action objective.

The paper specifically conjectures NP-hardness and notes that scheduling contributes an additional optimization dimension beyond classical bandwidth or layout problems.