We explore opportunities for unconventional information processing systems enabled by emerging photonic integration platforms.
Neuromorphic photonics combines photonic device physics with distributed processing models, resulting in a new class of ultrafast information processors. Key research includes laying out neuromorphic systems, packaging and controlling them, and demonstrating information processing functions.
Cryogenic silicon photonics brings together a new set of superconducting and optoelectronic phenomena with applications to integrated quantum information science. A particular focus is on silicon light sources. Silicon does not emit light at room temperature. Silicon light sources could be a next step in fully monolithic optoelectronic systems.
Neuromorphic hardware cannot implement most conventional algorithms. Harnessing their potential performance relies on developing algorithms that 1) map to the hardware, 2) outperform conventional approaches, and 3) solve tasks that are relevant to applications. The most impactful applications of any new technology are those that are currently unaddressed, and interdisciplinary engineering research is needed to pinpoint them.