Related papers: Thermodynamic Limits of Physical Intelligence
Transmitting energy and information are two essential aspects of nature. Recent findings suggest they are closely related, while a quantitative equivalence between them is still unknown. This thus motivates us to ask: Can information…
The growing demand for data center capacity, driven by the growth of high-performance computing, cloud computing, and especially artificial intelligence, has led to a sharp increase in data center energy consumption. To improve energy…
Chiplet-based integration enables large-scale systems that combine diverse technologies, enabling higher yield, lower costs, and scalability, making them well-suited to AI workloads. Processing-in-Memory (PIM) has emerged as a promising…
Situations where a spontaneous process of energy or matter transfer is enhanced by an external device are widespread in nature (human sweating system, enzyme catalysis, facilitated diffusion across bio-membranes, industrial heat…
A thermodynamic formalism describing the efficiency of information learning is proposed, which is applicable for stochastic thermodynamic systems with multiple internal degree of freedom. The learning rate, entropy production rate (EPR),…
We study the thermodynamic cost associated with the erasure of one bit of information over a finite amount of time. We present a general framework for minimizing the average work required when full control of a system's microstates is…
Efficiently harvesting thermodynamic resources requires a precise understanding of their structure. This becomes explicit through the lens of information engines -- thermodynamic engines that use information as fuel. Maximizing the work…
The ever increasing number and complexity of energy-bound devices (such as the ones used in Internet of Things applications, smart phones, and mission critical systems) pose an important challenge on techniques to optimize their energy…
We study the optimal exergy efficiency and power for thermodynamic systems with Onsager-type "current-force" relationship describing the linear-response to external influences. We derive, in simple analytic forms, the maximum efficiency and…
Biological systems have to build models from their sensory data that allow them to efficiently process previously unseen inputs. Here, we study a neural network learning a linearly separable rule using examples provided by a teacher. We…
Industrial cyber physical systems operate under heterogeneous sensing, stochastic dynamics, and shifting process conditions, producing data that are often incomplete, unlabeled, imbalanced, and domain shifted. High-fidelity datasets remain…
Within an inherently classical perspective, there is always an unavoidable energy cost associated with the information deletion and this common lore is at the heart of the Landauer's conjecture that does not impose, per se, any relevant…
As AI systems scale to multi-chiplet and wafer-level architectures, the demand for ultra-high bandwidth and system scalability has outpaced the capabilities of electrical interconnects and computing units. Large-scale heterogeneous…
Federated Learning (FL) enables collaborative model training across distributed clients while preserving data privacy. However, optimizing both energy efficiency and model accuracy remains a challenge, given device and data heterogeneity.…
A unified thermodynamic formalism describing the efficiency of learning is proposed. First, we derive an inequality, which is more strength than Clausius's inequality, revealing the lower bound of the entropy-production rate of a subsystem.…
A restricted form of Landauer's Principle, independent of computational considerations, is shown to hold for thermal systems by reference to the joint entropy associated with conjugate observables. It is shown that the source of the…
Understanding how much energy is needed and dissipated as heat for a given computational system and for a given program is a physically interesting and practically important problem. However, the thermodynamic costs of computational systems…
The construction of models from data is a significant contributor to the energetic costs of computation. Because of this, understanding how foundational thermodynamic bounds apply to modeling algorithms will be increasingly important. Here,…
Motivated by recent experiments [Lee et al. Nature 498, 209 (2013)], we present here a detailed theoretical analysis of the Joule heating in current-carrying single-molecule junctions. By combining the Landauer approach for quantum…
Physics-informed neural network architectures have emerged as a powerful tool for developing flexible PDE solvers which easily assimilate data, but face challenges related to the PDE discretization underpinning them. By instead adapting a…