A monolithic 3D IGZO-RRAM-SRAM-integrated architecture for robust and efficient compute-in-memory enabling equivalent-ideal device metrics  

作　　者：Shengzhe YAN Zhaori CONG Zi WANG Zhuoyu DAI Zeyu GUO Zhihang QIAN Xufan LI Xu ZHENG Chuanke CHEN Nianduan LU Chunmeng DOU Guanhua YANG Xiaoxin XU Di GENG Jinshan YUE Lingfei WANG Ling LI Ming LIU 

机构地区：[1]Institute of Microelectronics of the Chinese Academy of Sciences,Beijing 100029,China [2]University of Chinese Academy of Sciences,Beijing 100049,China

出　　处：《Science China(Information Sciences)》2025年第2期318-333,共16页中国科学(信息科学)(英文版)

基　　金：supported in part by National Key R&D Program of China(Grant No.2022YFB3606902);National Natural Science Foundation of China(Grant Nos.62204256,92264204,62274178,62488201);Beijing Nova Program(Grant No.Z211100002121125);China Postdoctoral Science Foundation(Grant Nos.BX20220330,2021M703444);Strategic Priority Research Program of Chinese Academy of Sciences(Grant No.XDB44000000)。

摘　　要：Compute-in-memory(CIM)based on various devices such as static random access memory(SRAM)and resistive random access memory(RRAM)and other emerging devices such as indium-gallium-zinc-oxide(IGZO)transistor,magnetoresistive RAM(MRAM),and ferroelectric RAM(FeRAM)has been explored for better performance and energy efficiency on neural networks applications.However,CIM based on a single-type device suffers a variety of non-ideal metrics to reach the simultaneous optimal accuracy,density,and energy efficiency.This work presents equivalent-ideal CIM(Eq-CIM),a monolithic 3D(M3D)IGZO-RRAM-SRAM integrated architecture for robust and efficient compute-in-memory enabling equivalent-ideal device metrics.To overcome the non-ideality(variation,endurance,temperature,etc.)of the single-type devices,system-technology co-optimization(STCO)is performed.This work highlights the non-ideality-aware functionality breakdown for robust high accuracy and simultaneous high density/efficiency,by utilizing 2T0C IGZO for temporal activation storage,RRAM for high-density weight storage,and SRAM for accurate CIM.Device-to-algorithm variation transfer is applied to analyze the system-level accuracy.We benchmark Eq-CIM architecture on CIFAR-10/ImageNet,with 5.06×storage density and 5.05×/2.45×area/energy efficiency compared with single-type-device-based CIM,and high robustness(<0.27%accuracy loss)from-40℃to 120℃.

关 键 词：compute-in-memory(CIM) monolithic 3D RRAM IGZO 3D-stack simulation 

分 类 号：TP333[自动化与计算机技术—计算机系统结构]