Development of a Photonic Field-Programmable Gate Array (pFGA) for Software-Controlled Photonics

Abstract

Field programmable gate arrays (FPGA) have revolutionized many areas of electronics, from prototyping of dedicated circuits to machine learning accelerators. Here, we propose to develop highly customizable photonic FPGAs, which promise to have a similar impact on the field of photonics. At the core of the proposed pFPGA is a fully packaged module capable of implementing arbitrary 20x20 linear optical transformations. This proposed system is far larger than the state of the art and will address a wide range of uses in basic research and in practical applications. We will develop control interface and algorithms to enable programmable circuits with a fidelity exceeding 99.9%, as well as long-term stability enabled by in-situ feedback. Another pillar of the program is to make the pFPGA available to the AFOSR-relevant research activities to accelerate interest in and innovation in programmable photonics. At MIT, the program will conduct a series of proof-of-principle studies, including benchmarking new types of error correction schemes, studying the ultimate accuracy limit posed by fabrication constraints, and demonstrating large-scale optical neural networks capable of conducting forward inference at optical speeds. These applications will enable essential fundamental research, including error propagation in coherent processors and examining the ultimate limits of hybrid reversible-irreversible computing systems.

Document Details

Document Type
DoD Grant Award
Publication Date
Aug 12, 2021
Source ID
FA95502010113

Entities

People

  • Dirk Englund

Organizations

  • Air Force Office of Scientific Research
  • Massachusetts Institute of Technology
  • United States Air Force

Tags

Fields of Study

  • Physics

Readers

  • Distributed Systems and Data Platform Development
  • Integrated Circuit Design and Technology.
  • Quantum Dot Semiconductor Device Photonics and Graphene Optoelectronic Materials and THz Physics.

Technology Areas

  • AI & ML
  • AI & ML - DoD AI Strategy
  • Microelectronics