Affiliation:
1. Aristotle University of Thessaloniki
2. University of Ioannina
3. Eindhoven University of Technology
4. University Research Center of Ioannina (URCI)
Abstract
We review different technologies and architectures for neuromorphic photonic accelerators, spanning from bulk optics to photonic-integrated-circuits (PICs), and assess compute efficiency in OPs/Watt through the lens of a comparative study where key technology aspects are analyzed. With an emphasis on PIC neuromorphic accelerators, we shed light onto the latest advances in photonic and plasmonic modulation technologies for the realization of weighting elements in training and inference applications, and present a recently introduced scalable coherent crossbar layout. Finally, we stress that current technologies face challenges endowing photonic accelerators with compute efficiencies in the PetaOPs/W, and discuss future implementation pathways towards improving performance.
Funder
Hellenic Foundation for Research and Innovation
H2020 LEIT Information and Communication Technologies
Subject
Electronic, Optical and Magnetic Materials
Cited by
15 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献