发表机构
Washington State University（华盛顿州立大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究针对基于非易失性内存的PIM架构存在热噪声影响推理准确性的问题，提出ThRIve方法，利用低秩自适应实现热鲁棒推理，在异构PIM架构上实验，准确率与SRAM相当，且降低了能量延迟积。
AI 中文摘要
内存处理（PIM）已成为加速机器学习工作负载的一项有前途的技术。基于非易失性内存的PIM架构由于能够执行节能的矩阵向量乘法运算，实现了有效的机器学习加速。然而，这些设备存在诸如热噪声等不理想因素。这种噪声会改变与实际模型权重相对应的存储在内存单元中的值，从而损害推理准确性。在这项工作中，我们引入了ThRIve，一种噪声感知训练方法，它利用低秩自适应在异构PIM架构上实现热鲁棒推理。ThRIve在对热噪声不太敏感的硬件上选择性地存储这些低秩噪声感知参数，从而实现对温度引起的噪声变化的鲁棒性。ThRIve减轻了热噪声的影响，并防止在整个工作温度范围内推理准确性下降。实验结果表明，采用ThRIve的架构保持一致的推理准确性，平均准确率保持在理想（即无噪声）准确率的2%以内，并且在整个工作温度范围内准确率的变化保持在平均值的2%以内。所提出的方法实现了与基于热弹性静态随机存取存储器（SRAM）的PIM系统相当的准确性和鲁棒性，同时在CNN模型推理期间将能量延迟积（EDP）降低了5.4倍。
英文摘要
Processing-In-Memory (PIM) has emerged as a promising technology for accelerating machine learning (ML) workloads. Specifically, non-volatile memory-based PIM architectures have enabled effective ML acceleration due to their ability to perform energy-efficient matrix-vector multiplication operations. However, these devices suffer from non-idealities such as thermal noise. This noise alters the stored values in the memory cells which correspond to actual model weights, compromising the inference accuracy. In this work, we introduce ThRIve, a noise-aware training methodology that leverages low-rank adaptation to enable thermally robust inference on heterogeneous PIM architectures. ThRIve selectively stores these low-rank noise-aware parameters on a hardware that is less susceptible to thermal noise, enabling robustness against temperature-induced noise variations. ThRIve mitigates the effects of thermal-noise and prevent the drop in inference accuracy across the entire operating temperature range. Experimental results demonstrate that ThRIve-enabled architectures maintain consistent inference accuracy, with the mean accuracy staying within 2% of the ideal (i.e., noise-free) accuracy, and the variation in accuracy across the entire operating temperature range remaining within 2% of the mean. The proposed methodology achieves accuracy and robustness comparable to thermally-resilient Static Random-Access Memory (SRAM)-based PIM systems, while delivering up to 5.4x reduction in energy-delay product (EDP) during CNN model inferencing.
CommentsPublished in ACM Transactions on Design Automation of Electronic Systems, Volume 31, Issue 2
Journal refACM Trans. Des. Autom. Electron. Syst. 31, 2, Article 23 (March 2026), 25 pages
DOI：10.1145/3774328
