[Invited Paper] Time-Ordered Computation (τ-QC) for Low-Power Continuous Information Processing: Applications to Generative Video and Medical Wearables

This paper introduces a design framework in which time is treated as an internal ordering structure, represented by a discrete progression index τ that governs the order of state generation. Building on this viewpoint, we present τ-QC (time-ordered computation), an engineering architecture that organizes computation as τ-ordered state generation and consistency checking. Here, “quantum” is used only as structural inspiration (“quantum-morphed” formalism), not as a claim about physical qubits, entanglement, or quantum computing hardware. The framework supports deterministic, low-power, continuous information processing by reallocating computation from stochastic sampling and heavy numerical pipelines to structured planning and verification. We illustrate the engineering implications through two case studies: (i) τ-based generative video, which shifts from diffusion-first refinement to symbolic-first generation with bounded diffusion only where needed, enabling smoother scaling with temporal length; and (ii) Rosie SG, a wearable inference profile that applies τ-ordered state verification to continuous physiological signals for early risk prediction under strict constraints on power, latency, and privacy. Overall, this work provides a unified architectural perspective centered on τ for rethinking time-structured computation across media generation, healthcare, and edge intelligence.