Brownian Computation

Brownian computers (or automata) are a form of abstract, theoretical computing devices. They are thought experiments rather than practical machines, but they are still based on, and derived from, physical laws. Their behavior is closely tied to principles involving entropy and Brownian motion. Due to their nature,they relate to many fields and branches, including (computational) thermodynamics, information theory, and Hamiltonian dynamics. With some ideas even extending into philosophy.

In theory, they have been proposed as a type of thermodynamically reversible computer, which would exempt them from Landauer’s principle; however, this remains under significant academic debate, as many argue that they are actually thermodynamically irreversible.

History

Brownian computers were first proposed as a solution to "Maxwell's daemon" by Charles H. Bennett (alongside contributions from Rolf Landauer). They took heavy inspiration from nature, and biological processes - namely DNA and RNA interactions (Nucleic Acid Interactions). The idea was developed in the context of IBM research and academic discussions in the fields of the thermodynamics of computation and theoretical computer science. Dr Bennett explored these ideas in the 1970s and 1980s, with key publications including Bennett’s 1973 paper “Logical Reversibility of Computation” and later works with Landauer in 1985. These publications shared the common goal of attempting to circumvent Landauer’s principle.

Mechanism

All Brownian computers are based on common principles aimed at minimizing or mitigating entropy production. They do not contain springs or sources of friction, as these would dissipate heat and defeat the objective. They attempt to be thermodynamically reversible by relying solely on logically reversible operations (e.g., NOT, Controlled NOT, etc.). Finally, they operate via thermal fluctuations—Brownian motion. These principles that define their mechanism are notably abstract and somewhat superficial due to the theoretical nature of the automata and the fact that they come in many different types.

Feynman lectures on computation

Richard Phillips Feynman, conventionally known, and recognized (Joint Nobel Prize 1965 - Physics) for his contributions to theoretical physics, quantum electrodynamics, and his renowned pedagogy (I.e. Feynman lectures on physics). However he was also an early computer scientist, having worked with and programmed early IBM computers to solve complex physics problems during and for the Manhattan project. Focusing primarily on its theoretical aspects, and niche cross overs with physics. Thus he was intrigued by the concept of Brownian Computers. He made heavy reference (alludes primarily, names occasionally) to them on a conceptual bases in his 1983-1986 series of lectures on "Potentialities and Limitations of Computing Machines" (in the Thermodynamics Section). Likewise to his - more well known - lectures on physics, these lectures were recorded, transcribed and then postmortemly compiled into a book in honor of Dr Feynman by fellows at caltech university. The book in question can generally be found by the name of “Feynman lectures on computation”.

References

This article "Brownian Computation" is from Wikipedia. The list of its authors can be seen in its historical and/or the page Edithistory:Brownian Computation. Articles copied from Draft Namespace on Wikipedia could be seen on the Draft Namespace of Wikipedia and not main one.

---