SinoGreenTech Academic Portal
Official PDF TranslationSCIENCE CHINA Materials

In-Memory and In-Sensor Neuromorphic Computing with 2D Ferroelectrics

Authors: ZHANG Qirui; CAO Guiming; PAN Er; YANG Fan; WANG Xuemei; CHEN Jiangang; WEN Zhixing; LIU Qing; LUO Xiao; LIU Fucai

DOI: 10.1007/s40843-026-4479-8Status: Verified Translated Edition
Sponsored AdvertisementAd Placement Area
reCAPTCHA Bot Shield Active

Preparing Secure Academic Download

Verifying human reader & generating high-resolution document...

Verifying Document Integrity15s remaining
← Back to Article
Protected by Google reCAPTCHA v3.PrivacyTerms
Sponsored ContentAdSense In-Feed Ad Slot

Key Findings in This Report

• • 2D ferroelectrics exhibit intrinsic non-volatility and atomic-scale thickness, enabling ultra-low power consumption and high fatigue endurance; these properties directly address the von Neumann bottleneck's energy and data-transfer inefficiencies, critical for edge AI and IoT devices where battery life and thermal budgets are constrained. • • Ferroelectric tunnel junctions, diodes, transistors, and photovoltaic devices are identified as key architectures; their integration into in-memory and in-sensor computing allows for artificial neural networks, spiking neural networks, and reservoir computing, reducing latency and energy overhead associated with frequent memory-processor data shuttling. • • The dangling-bond-free surfaces of 2D ferroelectrics facilitate heterogeneous integration with other materials, potentially enabling monolithic 3D integration of sensing, memory, and computing units; this is essential for scaling neuromorphic hardware beyond current CMOS limitations. • • Neuromorphic perception using 2D ferroelectrics offers efficient information processing and intelligent sensing, with potential applications in autonomous driving and real-time edge analytics; however, industrial adoption requires demonstration of wafer-scale uniformity, endurance beyond 10^12 cycles, and retention exceeding 10 years at elevated temperatures.