Christopher Bennett

Christopher is a 3rd year PhD student who is studying feasible applications of artificial synapses at the nano-scale (nano-synapses),  in particular , their application to learning tasks when combined in small hardware sub-systems. Studied tasks include simple ones such as elementary logic gate emulation, and anomaly detection, to more complex ones considered in such  fields as machine learning (multi-class classification). So far, his work has focused mostly on application and integration of a variety of physics-based models to larger scale systems, including notably work building on top of unique polymeric non-volatile nanosynapses (‘organic memristors’), as well as silver ionic  electrochemical cell devices. Particular focuses of his work have been on exploiting natural features of these devices, such as variability, and time parameters of the device , as a benefit rather than detriment to small on-chip learning systems, and exploring a variety of possible algorithms from simple to complex to evaluate their fragility. The goal of his thesis will finally be to contrast a variety of feasible on-chip algorithms and architectures, in terms of energy, accuracy, and overhead (area of system) constraints, with the ultimate goal of bringing new ideas from nanoelectronics to  design new hardware systems that can accelerate computing at a variety of scales.


Phone: (+33) 1 69 15 37 04