Rescuing RRAM-Based Computing From Static and Dynamic Faults
Rescuing RRAM-Based Computing From Static and Dynamic Faults
复制标题
DOI：
10.1109/tcad.2020.3037316
复制
发表时间：
2021-10-01
影响因子：
2.9
通讯作者：
Xie, Yuan
中科院分区：
文献类型：
--
作者：
Lin, Jilan;Wen, Cheng-Da;Xie, Yuan
新兴的电阻随机访问记忆（RRAM）显示了内存处理能力的巨大潜力，因此在加速记忆密集型应用程序（例如神经网络（NNS））方面吸引了相当大的研究兴趣。但是，由于RRAM细胞的耐药性的固有统计变化，基于RRAM的NN计算的准确性可以显着降解。在本文中，我们提出了Sight，这是一种协同的算法 - 体系结构耐受耐受性的框架，以整体解决此问题。具体而言，我们考虑了RRAM计算的三种主要故障：1）非线性电阻分布； 2）静态变化； 3）动态变化。从算法水平开始，我们提出了一种电阻感知的量化，以迫使NN参数遵循确切的非线性电阻分布作为RRAM，并引入输入调节技术以补偿RRAM变化。我们还提出了一个选择性的重量刷新方案，以解决运行时发生的动态变化问题。从体系结构层面，我们提出了一个普通和低成本的体系结构，以支持我们的容忍度耐受性方案。我们的评估表明，我们的三种容忍算法几乎没有准确的损失，并且拟议的视力体系结构使开销的性能低至7.14％。
Emerging resistive random access memory (RRAM) has shown the great potential of in-memory processing capability, and thus attracts considerable research interests in accelerating memory-intensive applications, such as neural networks (NNs). However, the accuracy of RRAM-based NN computing can degrade significantly, due to the intrinsic statistical variations of the resistance of RRAM cells. In this article, we propose SIGHT, a synergistic algorithm-architecture fault-tolerant framework, to holistically address this issue. Specifically, we consider three major types of faults for RRAM computing: 1) nonlinear resistance distribution; 2) static variation; and 3) dynamic variation. From the algorithm level, we propose a resistance-aware quantization to compel the NN parameters to follow the exact nonlinear resistance distribution as RRAM, and introduce an input regulation technique to compensate for RRAM variations. We also propose a selective weight refreshing scheme to address the dynamic variation issue that occurs at runtime. From the architecture level, we propose a general and low-cost architecture accordingly for supporting our fault-tolerant scheme. Our evaluation demonstrates almost no accuracy loss for our three fault-tolerant algorithms, and the proposed SIGHT architecture incurs performance overhead as little as 7.14%.