Paper Info
Title
Enabling High-Quality Uncertainty Quantification in a PIM Designed for Bayesian Neural Network
Abstract
Uncertainty quantification measures the prediction uncertainty of a neural network facing out-of-training-distribution samples. Bayesian Neural Networks (BNNs) can provide high-quality uncertainty quantification by introducing specific noise to the weights during inference. To accelerate BNN inference, ReRAM processing-in-memory (PIM) architecture is a competitive solution to provide both high-efficient computing and in-situ noise generation at the same time. However, there normally exists a huge gap between the generated noise in PIM hardware and that required by a BNN model. We demonstrate that the quality of uncertainty quantification is substantially degraded due to this gap. To solve this problem, we propose a holistic framework called W2W-PIM. We first introduce an efficient method to generate noise in ReRAM PIM design according to the demand of a BNN model. In addition, the PIM architecture is carefully modified to enable the noise generation and evaluate uncertainty quality. Moreover, a calibration unit is further introduced to reduce the noise gap caused by imperfection of the noise model. Comprehensive evaluation results demonstrate that W2W-PIM framework can achieve high-quality uncertainty quantification and high energy-efficiency at the same time.
Year
DOI
Venue
2022
10.1109/HPCA53966.2022.00080
2022 IEEE International Symposium on High-Performance Computer Architecture (HPCA)
Keywords
DocType
ISSN
ReRAM,Bayesian Neural Network,Analog Computing,Noise
Conference
1530-0897
ISBN
Citations 
PageRank 
978-1-6654-2028-0
0
0.34
References 
Authors
0
14
Name
Order
Citations
PageRank
Xingchen Li131.83
Bingzhe Wu2186.41
Guangyu Sun31920111.55
Zhe Zhang469.60
Zhihang Yuan500.34
Runsheng Wang616921.11
Ru Huang718848.74
Dimin Niu800.34
Hongzhong Zheng900.34
Zhichao Lu1000.34
Liang Zhao1100.34
Meng-Fan Chang1245945.63
Tianchan Guan1300.34
Xin Si1400.34