Freely scalable and reconfigurable optical hardware for deep learning
Abstract
As deep neural network (DNN) models grow ever-larger, they can achieve higher accuracy and solve more complex problems. This trend has been enabled by an increase in available compute power; however, efforts to continue to scale electronic processors are impeded by the costs of communication, thermal management, power delivery and clocking. To improve scalability, we propose a digital optical neural network (DONN) with intralayer optical interconnects and reconfigurable input values. The path-length-independence of optical energy consumption enables information locality between a transmitter and a large number of arbitrarily arranged receivers, which allows greater flexibility in architecture design to circumvent scaling limitations. In a proof-of-concept experiment, we demonstrate optical multicast in the classification of 500 MNIST images with a 3-layer, fully-connected network. We also analyze the energy consumption of the DONN and find that digital optical data transfer is beneficial over electronics when the spacing of computational units is on the order of $$>10\,\upmu $$ > 10 μ m.
Document Details
- Document Type
- Pub Defense Publication
- Publication Date
- Feb 04, 2021
- Source ID
- 10.1038/s41598-021-82543-3
Entities
People
- Alexander Sludds
- Dirk Englund
- Joel Emer
- Liane Bernstein
- Ryan Hamerly
- Vivienne Sze
Organizations
- Army Research Office
- NTT, Inc.
- National Science Foundation
- Natural Sciences and Engineering Research Council
- Oak Ridge Institute for Science and Education