The Thermodynamics of Natural and Artificial Distributed Computational Systems

Navigation Home Agenda Participants Background Papers

Organizers

David Wolpert (SFI), Christopher Lynn (Princeton), Joshua Grochow (U.C. Boulder; SFI), Jan Korbel (CSH Vienna), and Farita Tasnim (MIT)

Workshop Dates

August 15-17, 2022

---

Description

The thermodynamics of computation is a long-standing interest in the physics, computer science, and biology communities, playing a major role in issues ranging from the design of artificial digital systems to the foundations of physics to theoretical neurobiology. The revolution in non-equilibrium statistical physics of the past two decades, sometimes summarized as "stochastic thermodynamics", has provided a major advance in our ability to investigate this topic. Such research will be the focus of this summer's workshop. Specifically, we will focus on systems that:

1. Are distributed, with multiple spatially separated subsystems;
2. Are not at thermodynamic equilibrium (and in general, not even in a stationary state);
3. Have substantial thermodynamic costs of communication among the subsystems and substantial thermodynamic costs of the information processing within the subsystems.

Computational systems with these three characteristics are ubiquitous, ranging from digital devices to brains to genetic circuits. Moreover, we have lots of evidence that thermodynamic costs have played a major role in determining the physical architecture of such systems. Yet to date, there has been almost no application of stochastic thermodynamics to investigate these systems in order to deepen our understanding of the relationship among their energetic behavior, computational behavior, robustness, etc. Our conference aims to identify challenges, opportunities, and research priorities to push this research forward.