Les ordinateurs thermodynamiques suivent le flux de l'énergie

Researchers are exploring a novel approach to computing that leverages the chaotic motion of atoms caused by heat, typically viewed as a disruptive force, to perform computations efficiently. This emerging field, known as thermodynamic computing, aims to harness thermal fluctuations rather than suppress them, potentially leading to devices that require significantly less power and generate minimal heat. Recent simulations using conventional silicon-based logic circuits suggest that the foundational principles of thermodynamic computing may be viable. The concept builds upon the second law of thermodynamics, which states that the entropy of a closed system tends to increase over time, resulting in greater disorder. While this often leads to energy being lost as random thermal fluctuations, certain natural processes utilize these fluctuations to achieve more structured outcomes. For instance, proteins fold into specific three-dimensional shapes due to thermal movements, enabling biological functions essential for life. Similarly, thermodynamic computing proposes that systems can be designed to use these inherent fluctuations as a computational resource. Patrick Coles, a physicist at the startup Normal Computing in New York, explained that the field focuses on creating computers that actively employ thermodynamics as part of their operational framework. If successful, this could revolutionize both the computing industry and our understanding of computation itself. Unlike traditional computing methods that strive to minimize interference from environmental noise, thermodynamic computing seeks to integrate and benefit from such disturbances. David Sivak, a statistical physicist at Simon Fraser University in Burnaby, Canada, noted that the idea behind thermodynamic computing involves leveraging the computations that occur naturally within the environment. These processes, though not traditionally recognized as computational, involve systems navigating through an "energy landscape," a conceptual map representing the varying energies associated with different configurations of a system's components. In this context, lower energy states correspond to stable configurations, whereas higher energy states represent less stable conditions. To illustrate this concept, consider the digestion of milk. Lactase, an enzyme crucial for breaking down lactose, achieves its functional form through a process influenced by thermal fluctuations. As the protein chain forms, these fluctuations enable it to explore various configurations until it stabilizes in its most efficient structure. This natural process mirrors the principles underlying thermodynamic computing, where systems navigate through energy landscapes to reach optimal states. Recent advancements in simulating thermodynamic computation within existing silicon-based technology indicate that the theoretical underpinnings of this approach hold practical potential. By utilizing the intrinsic properties of materials and the natural tendencies of systems to seek stability, researchers aim to develop new types of computing architectures that operate with unprecedented efficiency. As the field continues to evolve, scientists are investigating how best to implement these principles in real-world applications. The challenge lies in translating the theoretical models into tangible technologies capable of performing meaningful computations. Despite the promising initial results, much work remains to fully realize the vision of thermodynamic computing. Researchers are optimistic that continued exploration will yield breakthroughs that redefine the boundaries of modern computing.