Continuum Free-Energy Computing

TL;DR

This paper proposes a novel computing paradigm called continuum free-energy computing, which encodes problems in free-energy functionals and solves them via relaxational dynamics, with a potential physical realization using ion-patterned FeRh.

Contribution

It introduces the concept of continuum free-energy computing, linking physical free-energy landscapes to computational processes and identifying a possible material implementation.

Findings

Proposes a new computing paradigm based on free-energy landscapes.

Identifies ion-patterned FeRh as a potential physical realization.

Defines task classes, protocols, and physical constraints for this paradigm.

Abstract

Building on nonintrinsic Landau theory, we introduce continuum free-energy computing as a new computing paradigm in which problem instances are encoded in programmable free-energy functionals and solved by intrinsic relaxational dynamics. We identify ion-patterned FeRh as a plausible physical realization through spatial control of the local phase bias, with antiferromagnetic-ferromagnetic interface motion providing the relaxational mechanism. We further identify two representative task classes, a minimal operating protocol, and the main physical constraints.