Near-energy-free photonic Fourier transformation for convolution operation acceleration

By Hangbo Yang ^a,b, Nicola Peserico ^a,b, Shurui Li ^c, Xiaoxuan Ma ^d, Russell L. T. Schwartz ^a,b, Mostafa Hosseini ^c, Aydin Babakhani ^c, Chee Wei Wong ^c, Puneet Gupta ^c and Volker J. Sorger ^a,b 
a University of Florida, Department of Electrical and Computer Engineering, Gainesville, Florida, United States
^b University of Florida, Florida Semiconductor Institute, Gainesville, Florida, United States
^c University of California Los Angeles, Department of Electrical and Computer Engineering, Los Angeles, California, United States
^d The George Washington University, Department of Electrical and Computer Engineering, Washington, DC, United States

Abstract

Convolutional operations are computationally intensive in artificial intelligence (AI) services, and their overhead in electronic hardware limits machine learning scaling. Here, we introduce a photonic joint transform correlator (pJTC) using a near-energy-free on-chip Fourier transformation to accelerate convolution operations. The pJTC reduces computational complexity for both convolution and cross-correlation from O ( N⁴ ) to O ( N² ) , where N² is the input data size. Demonstrating functional Fourier transforms and convolution, this pJTC achieves 98.0% accuracy on an exemplary Modified National Institute of Standards and Technology inference task. Furthermore, a wavelength-multiplexed pJTC architecture shows potential for high throughput and energy efficiency, reaching 305 TOPS / W and 40.2 TOPS / mm², based on currently available foundry processes. An efficient, compact, and low-latency convolution accelerator promises to advance next-generation AI capabilities across edge demands, high-performance computing, and cloud services.

To read the full article, click here