The Decision Loop as a Hidden Limit of the Effective Speed of Information Transfer

Contemporary computational architectures are traditionally optimized in terms of energy consumption, clock frequency, and parallelism. However, this approach systematically neglects the physical costs associated with decision-making, measurement, and irreversible information discarding. This preprint shows that the true universal limitation of modern computational performance is neither signal propagation speed nor the intrinsic evolution of computational states, but the existence of a critical decision loop in which irreversible decisions occur. Within the framework of Information–Energetic Thermodynamics (IET), a general design methodology is formulated that quantifies energetic, temporal, and informational costs of decision processes in both classical and quantum systems. The core finding is the IET–Decision Bound, a physical limit on system performance arising from the rate and cost of irreversible decisions. Architectural strategies such as deferred verification, decision batching, and speculative internal states are proposed to mitigate these bottlenecks. The framework yields testable predictions and provides measurable guidelines for improving performance in both classical and quantum systems by minimizing decision-induced overhead. Keywords: Decision loop, IET–Decision Bound, computational bottlenecks, information thermodynamics, quantum measurement, control overhead, architectural design