Measured hardware and higher-order physical realization

Establish measured closed-loop hardware implementations, higher-order physical realizations, and validated latency and energy characteristics for the Radio Alternating Operator Ansatz beyond simulation-level evidence.

Background

The paper evaluates programmable radio propagation through differentiable software models and constrained passive phase-only free-space simulations rather than a physical RF prototype. The reported propagation results therefore do not establish calibrated hardware behavior, measured closed-loop operation, latency, or energy consumption.

The RAOA experiments also represent native third- and fourth-order local fields directly in software. The phase-only propagation compiler is evaluated only for pairwise operators, leaving both higher-order physical realization and hardware validation unresolved.

References

The evidence remains operator- and simulation-level: higher-order realization, calibrated hardware, and measured latency and energy remain open.

— RAOA: Alternating-Operator Neural Computation with Programmable Radio Propagation  (2610.02683 - Koike-Akino, 2 Oct 2026) in Section 5, Discussion, paragraph “Scope”

The companion paper's own energy section states plainly that holding $P_i$ constant as problem size grows is ``a physically reasonable assumption... but it is exactly that: an assumption, not yet a validated one.''

— Closing the Prediction Gap: Completing Machine Shape So That Predicted Time, Power, Energy, and Mapping Match What Real Hardware Does  (2610.03197 - Mullin et al., 2 Oct 2026) in Section 5.2, “$P_d$: power draw, indexed by realization, not device alone”