Ultralow latency computation based on integrated nanophotonics

Masaya Notomi, Kengo Nozaki, Shota Kita, Akihiko Shinya, Tohru Ishihara, Inoue Koji

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Moore's law for CMOS computers is still continuing, but its near-future saturation is now being discussed. One of the serious saturations is about its latency. The computation delay for a CMOS transistor is already saturated above 10 ps, which will be problematic when ultralow-latency response is required for broad-band data streams, even with parallelization or pipe-line processing. We regard that optical circuits may serve as ultralow-latency computation circuits if they are small enough and tightly combined with electronic circuits. The former requires nanophotonic devices/circuits and the former requires OE/EO conversion with ultrasmall capacitance.

Original languageEnglish
Title of host publicationJSAP-OSA Joint Symposia, JSAP 2018
PublisherOSA - The Optical Society
ISBN (Print)9784863486942
Publication statusPublished - Jan 1 2018
EventJSAP-OSA Joint Symposia, JSAP 2018 - Nagoya, Japan
Duration: Sep 18 2018Sep 21 2018

Publication series

NameOptics InfoBase Conference Papers
VolumePart F125-JSAP 2018

Conference

ConferenceJSAP-OSA Joint Symposia, JSAP 2018
CountryJapan
CityNagoya
Period9/18/189/21/18

    Fingerprint

All Science Journal Classification (ASJC) codes

  • Electronic, Optical and Magnetic Materials
  • Mechanics of Materials

Cite this

Notomi, M., Nozaki, K., Kita, S., Shinya, A., Ishihara, T., & Koji, I. (2018). Ultralow latency computation based on integrated nanophotonics. In JSAP-OSA Joint Symposia, JSAP 2018 (Optics InfoBase Conference Papers; Vol. Part F125-JSAP 2018). OSA - The Optical Society.