Domain-specific neural network accelerators have seen growing interest in recent years due to their improved energy efficiency and inference performance compared to CPUs and GPUs. In this paper, we propose a novel cross-layer optimized neural network accelerator called CrossLight that leverages silicon photonics. CrossLight includes device-level engineering for resilience to process variations and thermal crosstalk, circuit-level tuning enhancements for inference latency reduction, and architecture-level optimization to enable higher resolution, better energy-efficiency, and improved throughput. On average, CrossLight offers 9.5x lower energy-per-bit and 15.9x higher performance-per-watt at 16-bit resolution than state-of-the-art photonic deep learning accelerators.
翻译:近些年来,由于与CPU和GPU相比能效和推断性能的提高,特定神经网络加速器受到越来越多的关注。在本文中,我们建议采用新型的跨层优化神经网络加速器,名为CrossLight,利用硅光子。交叉光线包括用于处理变异的抗御力设备级工程、热交叉跟踪、降低延缓率的电路级调控增强、以及结构级优化,以便实现更高的分辨率、更高的能效和更高的吞吐量。 平均而言,CrossLight提供比最新光学深层加速器16位分辨率低9.5x每位能量和高15.9x每瓦性能。