According to Landauer, information is physical. And erasing one bit of information costs at least kT ln(2) of energy. This fundamental limit to the energy efficiency of irreversible classical computing has been verified many times, but only in systems where energy and information are stored in matter. In this project, you will measure the minimum energy required for bit erasure in systems where energy and information are stored in light. Such optical systems have recently emerged as promising platforms for energy efficient computing. However, fundamental limits to information processing in such systems are largely unknown. To gain insight into this problem, you will perform experiments with a laser-driven nonlinear optical cavity working as a bit.  You will set and reset (i.e, erase) this bit by varying the laser power and/or the cavity length. Your main goal is to quantify and reach the fundamental limits to the energy efficiency, speed, and accuracy of resetting an optical bit. Moreover, you will develop methods to increase the optical nonlinearity of the system, so that quantum effects emerge and the bit becomes a qubit. Your ultimate goal is assess whether and how quantum effects modify fundamental limits to classical bit erasure. Your main focus will be on experiments, but you will closely collaborate with Christopher Jarzynski and a PhD student in his group at the University of Maryland. Together, you will develop a stochastic thermodynamic framework for optical systems across the semiclassical to quantum transition