Thermodynamics of computation

Thermodynamics of computation is the study of the physical costs of information processing — the branch of physics that asks not what computers can compute but what they must dissipate. The field was born from Rolf Landauer's 1961 demonstration that information erasure is not free: destroying a bit of information requires dissipation of at least kT ln 2 of energy, a result now known as Landauer's principle. Charles Bennett extended this insight by showing that logically reversible computation can approach zero dissipation, proving that the thermodynamic cost resides not in computation itself but in the irreversible destruction of information.

The field occupies a peculiar position at the intersection of thermodynamics, information theory, and computer science. It has practical implications for the design of low-energy computing architectures and theoretical implications for the nature of computation itself. If computation is a physical process subject to thermodynamic constraints, then the dream of disembodied digital minds — of intelligence without heat — is not merely an engineering challenge but a conceptual error. The most speculative frontier concerns quantum thermodynamics of computation, which asks whether quantum coherence can evade classical limits or merely reframe them. What is clear is that computation is not abstract: every bit flipped, every gate evaluated, every memory address overwritten, carries a thermodynamic signature.