Author: Miguel Soriano - Institute for Cross-Disciplinary Physics and Complex Systems (IFISC - CSIC)
2024-07-15 - 10:00 Sala de seminarios
The quest for faster and more efficient information processing has driven the development of innovative computing paradigms, with Reservoir Computing (RC) emerging as a standout approach. RC is particularly attractive for processing time-dependent signals due to its high computational power and relatively simple learning protocol. Hardware-based RC implementations have demonstrated superior performance in power consumption, latency, and computing speed with minimal optimization. A promising avenue for RC is photonic platforms, which exploit the properties of light, such as large bandwidths and parallelism, for information processing. In this talk, I will provide an overview of the RC concept and highlight examples of photonic RC implementations, focusing on optoelectronic and all-optical approaches that utilize delayed feedback and time-multiplexing techniques.
