O-HAS: Optical Hardware Accelerator Search for Boosting Both Acceleration Performance and Development Speed
O-HAS: Optical Hardware Accelerator Search for Boosting Both Acceleration Performance and Development Speed
复制标题
DOI：
10.1109/iccad51958.2021.9643442
复制
发表时间：
2021-08
期刊：
影响因子：
--
通讯作者：
Mengquan Li;Zhongzhi Yu;Yongan Zhang;Yonggan Fu;Yingyan Lin
中科院分区：
文献类型：
--
作者：
Mengquan Li;Zhongzhi Yu;Yongan Zhang;Yonggan Fu;Yingyan Lin
深度神经网络（DNN）最近的突破和令人望而却步的复杂性激发了对特定领域DNN加速器的广泛兴趣，其中光学DNN加速器特别有前途，这要归功于其实现上级性能/瓦的前所未有的潜力。然而，光学DNN加速器的发展比电DNN加速器慢得多。一个关键的挑战是，虽然已经开发了许多技术来促进电DNN加速器的开发，但支持或加速光学DNN加速器设计的技术仍然很少探索，限制了光学DNN加速器的可实现性能和创新发展。为此，我们开发了首个名为O-HAS的框架，该框架首次展示了自动化光学硬件加速器搜索，以提高光学DNN加速器的加速效率和开发速度。具体而言，我们的O-HAS由两个集成的使能器组成：（1）O-成本预测器，其可以基于DNN模型参数和光学加速器设计准确而有效地预测光学加速器的能量和延迟;以及（2）O-搜索引擎，其可以自动探索光学DNN加速器的大设计空间并识别最佳加速器（即，微体系结构和算法到加速器的映射方法），以便最大化目标加速效率。广泛的实验和烧蚀研究一致验证了我们的O-Cost Predictor和O-Search Engine的有效性以及O-HAS产生的光学加速器的出色效率。
The recent breakthroughs and prohibitive complexities of Deep Neural Networks (DNNs) have excited extensive interest in domain specific DNN accelerators, among which optical DNN accelerators are particularly promising thanks to their unprecedented potential of achieving superior performance-per-watt. However, the development of optical DNN accelerators is much slower than that of electrical DNN accelerators. One key challenge is that while many techniques have been developed to facilitate the development of electrical DNN accelerators, techniques that support or expedite optical DNN accelerator design remain much less explored, limiting both the achievable performance and the innovation development of optical DNN accelerators. To this end, we develop the first-of-its-kind framework dubbed O-HAS, which for the first time demonstrates automated Optical Hardware Accelerator Search for boosting both the acceleration efficiency and development speed of optical DNN accelerators. Specifically, our O-HAS consists of two integrated enablers: (1) an O-Cost Predictor, which can accurately yet efficiently predict an optical accelerator's energy and latency based on the DNN model parameters and the optical accelerator design; and (2) an O-Search Engine, which can automatically explore the large design space of optical DNN accelerators and identify the optimal accelerators (i.e., the micro-architectures and algorithm-to-accelerator mapping methods) in order to maximize the target acceleration efficiency. Extensive experiments and ablation studies consistently validate the effectiveness of both our O-Cost Predictor and O-Search Engine as well as the excellent efficiency of O-HAS generated optical accelerators.