期刊
IEEE TRANSACTIONS ON ELECTRON DEVICES
卷 67, 期 10, 页码 4185-4192出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TED.2020.3015178
关键词
Quantization (signal); Resistance; Random access memory; Hardware; Decoding; Parallel processing; Prototypes; Deep neural networks (DNNs); in-memory computing (IMC); monolithic integration; nonvolatile memory (NVM); resistive RAM (RRAM)
资金
- NSF-SRC-E2CDA [2018-NC-2762B]
- NSF [1652866, 1715443, 1740225]
- JUMP C-BRIC
- JUMP ASCENT (SRC Program - Defense Advanced Research Projects Agency (DARPA))
Deep neural network (DNN) hardware designs have been bottlenecked by conventional memories, such as SRAM due to density, leakage, and parallel computing challenges. Resistive devices can address the density and volatility issues but have been limited by peripheral circuit integration. In this work, we present a resistive RAM (RRAM)-based in-memory computing (IMC) design, which is fabricated in 90-nm CMOS with monolithic integration of RRAM devices. We integrated a 128 x 64 RRAM array with CMOS peripheral circuits, including row/column decoders and flash analog-to-digital converters (ADCs), which collectively become a core component for scalable RRAM-based IMC for large DNNs. To maximize IMC parallelism, we assert all 128 wordlines of the RRAM array simultaneously, perform analog computing along the bitlines, and digitize the bitline voltages using ADCs. The resistance distribution of low-resistance states is tightened by an iterative write-verify scheme. Prototype chip measurements demonstrate high binary DNN accuracy of 98.5% for MNIST and 83.5% for CIFAR-10 data sets, with 24 TOPS/W and 158 GOPS. This represents 22.3x and 10.1x improvements in throughput and energy-delay product (EDP), respectively, compared with the state-of-the-art literature, which can enable intelligent functionalities for area-/energy-constrained edge computing devices.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据