arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2004.00802cs.NE
收藏

SONOS非易失性存储器阵列中的设备感知推断操作

Device-aware inference operations in SONOS nonvolatile memory arrays

Christopher H. Bennett, T. Patrick Xiao, Ryan Dellana, Vineet Agrawal, Ben Feinberg, Venkatraman Prabhakar, Krishnaswamy Ramkumar, Long Hinh, Swatilekha Saha, V… 展开作者

Christopher H. Bennett, T. Patrick Xiao, Ryan Dellana, Vineet Agrawal, Ben Feinberg, Venkatraman Prabhakar, Krishnaswamy Ramkumar, Long Hinh, Swatilekha Saha, Vijay Raghavan, Ramesh Chettuvetty, Sapan Agarwal, Matthew J. Marinella

摘要:

非易失性存储器阵列可以部署预训练的神经网络模型用于边缘推断。然而,这些系统受到设备级噪声和保留问题的影响。在这里,我们研究这些影响所造成的损害,介绍一种缓解策略,并证明其在制造阵列的SONOS(硅-氧化物-氮化物-氧化物-硅)设备的使用。在MNIST,时尚MNIST和CIFAR-10任务中,我们的方法增加了对突触噪声和漂移的弹性。我们还表明,强大的性能可以实现与ADC的5-8位精度。

英文摘要:

Non-volatile memory arrays can deploy pre-trained neural network models for edge inference. However, these systems are affected by device-level noise and retention issues. Here, we examine damage caused by these effects, introduce a mitigation strategy, and demonstrate its use in fabricated array of SONOS (Silicon-Oxide-Nitride-Oxide-Silicon) devices. On MNIST, fashion-MNIST, and CIFAR-10 tasks, our approach increases resilience to synaptic noise and drift. We also show strong performance can be realized with ADCs of 5-8 bits precision.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑