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 14x better throughput and 8x better energy efficiency, in comparison to state-of-the-art transformer accelerators.

翻译：Transformer神经网络正在被迅速整合到自然语言处理（NLP）和计算机视觉的最先进解决方案中。然而，这些模型的复杂结构对于在传统电子平台上加速其执行造成了挑战。本文提出了第一个硅光子硬件神经网络加速器TRON，旨在加速基于Transformer的模型（如BERT和视觉Transformer）。我们的分析表明，与最先进的Transformer加速器相比，TRON的吞吐量至少提高了14倍，能效至少提高了8倍。