arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2006.05696cs.ARcs.ET
收藏

面向新兴NVM技术的统一表征平台:使用现成的NVM芯片进行神经网络应用基准测试

Unified Characterization Platform for Emerging NVM Technology: Neural Network Application Benchmarking Using off-the-shelf NVM Chips

Supriya Chakraborty, Abhishek Gupta, Manan Suri

摘要:

在本文中,我们提出了一个统一的基于FPGA的电气测试台,用于表征不同的新兴非易失性存储器(NVM)芯片。特别是,我们提出了详细的电气特性和基准测试的多个市售,现成的,NVM芯片,即:MRAM、FeRAM、CBRAM和ReRAM。我们调查重要的NVM参数,如:(i)电流消耗模式,(ii)耐久性,和(iii)错误特性。提出的基于FPGA的测试平台,然后利用概念验证(Proof-of-Concept,缩写)神经网络(NN)图像分类应用。四种新兴的NVM芯片在推理模式下作为主动权重存储器,以标准SRAM和闪存技术为基准,用于AI应用。

英文摘要:

In this paper, we present a unified FPGA based electrical test-bench for characterizing different emerging NonVolatile Memory (NVM) chips. In particular, we present detailed electrical characterization and benchmarking of multiple commercially available, off-the-shelf, NVM chips viz.: MRAM, FeRAM, CBRAM, and ReRAM. We investigate important NVM parameters such as: (i) current consumption patterns, (ii) endurance, and (iii) error characterization. The proposed FPGA based testbench is then utilized for a Proof-of-Concept (PoC) Neural Network (NN) image classification application. Four emerging NVM chips are benchmarked against standard SRAM and Flash technology for the AI application as active weight memory during inference mode.

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。

↑