Shift-BNN: Highly-Efficient Probabilistic Bayesian Neural Network Training via Memory-Friendly Pattern Retrieving
Shift-BNN: Highly-Efficient Probabilistic Bayesian Neural Network Training via Memory-Friendly Pattern Retrieving
复制标题
DOI：
10.1145/3466752.3480120
复制
发表时间：
2021-10
期刊：
影响因子：
--
通讯作者：
Qiyu Wan;Haojun Xia;Xingyao Zhang;Lening Wang;S. Song;Xin Fu
中科院分区：
文献类型：
--
作者：
Qiyu Wan;Haojun Xia;Xingyao Zhang;Lening Wang;S. Song;Xin Fu
拥有不确定性估计特性的贝叶斯神经网络（BNN）已在广泛的安全至关重要的AI应用中采用，这些应用要求可靠，可靠的决策，例如自动驾驶，救援机器人，医疗图像诊断概率BNN模型的过程涉及训练一个采样的DNN模型的集合，该模型诱发了比训练单个DNN模型的数量级数据移动的数量级。高斯随机变量（GRV）的片外数据传输。精度损失，在硬件级别上有效地支持我们的LFSR归还策略，我们探索当前DNN加速器的设计空间，并确定最佳的计算映射方案，以最大程度地适应我们的策略。有效的BNN训练加速器，名为Shift-BNN，在五个代表性的BNN模型上是低成本和可扩展的。在基线DNN训练加速器上的加速度高达2.8×）。
Bayesian Neural Networks (BNNs) that possess a property of uncertainty estimation have been increasingly adopted in a wide range of safety-critical AI applications which demand reliable and robust decision making, e.g., self-driving, rescue robots, medical image diagnosis. The training procedure of a probabilistic BNN model involves training an ensemble of sampled DNN models, which induces orders of magnitude larger volume of data movement than training a single DNN model. In this paper, we reveal that the root cause for BNN training inefficiency originates from the massive off-chip data transfer by Gaussian Random Variables (GRVs). To tackle this challenge, we propose a novel design that eliminates all the off-chip data transfer by GRVs through the reversed shifting of Linear Feedback Shift Registers (LFSRs) without incurring any training accuracy loss. To efficiently support our LFSR reversion strategy at the hardware level, we explore the design space of the current DNN accelerators and identify the optimal computation mapping scheme to best accommodate our strategy. By leveraging this finding, we design and prototype the first highly efficient BNN training accelerator, named Shift-BNN, that is low-cost and scalable. Extensive evaluation on five representative BNN models demonstrates that Shift-BNN achieves an average of 4.9 × (up to 10.8 ×) boost in energy efficiency and 1.6 × (up to 2.8 ×) speedup over the baseline DNN training accelerator.