一句话：提出 Explainable-DSE，用瓶颈分析引导深度学习加速器软硬件协同设计空间探索，显著减少迭代并降低延迟。
针对深度学习加速器软硬件协同设计空间巨大、现有黑盒 DSE 不分析设计低效而需大量试验的问题，提出 Explainable-DSE。该方法利用瓶颈模型提供的执行成本信息识别瓶颈，并据此在后续探索中采取缓解瓶颈的采集，从而可解释地推理设计低效。作者为 DNN 加速器构建瓶颈模型，并提出表达领域特定瓶颈模型并与 DSE 框架对接的 API。在近期计算机视觉和语言模型上评估表明，Explainable-DSE 主要探索有效候选，相比基于进化或机器学习的非可解释 DSE，以少 47 倍的迭代获得延迟低 6 倍的协同设计，仅需数分钟或数十次迭代，为运行时 DSE 提供机会。
Effective design space exploration (DSE) is paramount for hardware/software codesigns of deep learning accelerators that must meet strict execution constraints. For their vast search space, existing DSE techniques can require excessive trials to obtain a valid and efficient solution because they rely on black-box explorations that do not reason about design inefficiencies. In this paper, we propose Explainable-DSE - a framework for the DSE of accelerator codesigns using bottleneck analysis. By leveraging information about execution costs from bottleneck models, our DSE is able to identify bottlenecks and reason about design inefficiencies, thereby making bottleneck-mitigating acquisitions in further explorations. We describe the construction of bottleneck models for DNN accelerators. We also propose an API for expressing domain-specific bottleneck models and interfacing them with the DSE framework. Acquisitions of our DSE systematically cater to multiple bottlenecks that arise in executions of multi-functional workloads or multiple workloads with diverse execution characteristics. Evaluations for recent computer vision and language models show that Explainable-DSE mostly explores effectual candidates, achieving codesigns of 6X lower latency in 47X fewer iterations vs. non-explainable DSEs using evolutionary or ML-based optimizations. By taking minutes or tens of iterations, it enables opportunities for runtime DSEs.
@inproceedings{paperbot10126,
title = {Explainable-DSE: An Agile and Explainable Exploration of Efficient HW/SW Codesigns of Deep Learning Accelerators Using Bottleneck Analysis},
author = {Shail Dave and Tony Nowatzki and Aviral Shrivastava},
booktitle = {ASPLOS},
year = {2023},
doi = {10.1145/3623278.3624772}
}