TRON: transformer neural network acceleration with non-coherent silicon photonics

Abstract

Transformer neural networks are rapidly being integrated into state-of-the-art solutions for natural language processing (NLP) and computer vision. However, the complex structure of these models creates challenges for accelerating their execution on conventional electronic platforms. We propose the first silicon photonic hardware neural network accelerator called TRON for transformer-based models such as BERT, and Vision Transformers. Our analysis demonstrates that TRON exhibits at least 14× better throughput and 8× better energy efficiency, in comparison to state-of-the-art transformer accelerators.