What is it about?
This tutorial covers perspectives and design of photonic and optoelectronic neuromorphic computers, including insights on data encoding, network topologies, device architectures, and benchmarking. Advantages over electronic approaches and comparisons between existing devices are drawn.
Featured Image
Photo by Milad Fakurian on Unsplash
Why is it important?
Silicon photonic platforms are quickly growing in capability along with co-integrated CMOS and photonic platforms, which bring the high-bandwidth network advantages of wavelength-division multiplexing (WDM) to the new integrated circuit designs. Neuromorphic computers require large-scale interconnected networks between devices that is best served by this technology.
Perspectives
Neuromorphic computing is well-suited to many of the edge computing and autonomous processing tasks flooding the modern computing scene. As continued transistor scaling trends begin to slow, neuromorphic computing will likely emerge as a solution to continue improving computing speeds for highly-parllelized algorithms and tasks. Through this tutorial, I hope to convince the reader that photonic and optoelectronic solutions are worth further research and development efforts.
Luis El Srouji
University of California Davis
Read the Original
This page is a summary of: Photonic and optoelectronic neuromorphic computing, APL Photonics, May 2022, American Institute of Physics,
DOI: 10.1063/5.0072090.
You can read the full text:
Contributors
The following have contributed to this page
