TReCiM: Lower Power and Temperature-Resilient Multibit 2FeFET-1T Compute-in-Memory Design

2025-01-02

AI Summary

Subthreshold ferroelectric field-effect transistor (FeFET)-based in-memory computing suffers from accuracy degradation under temperature variations (0–85°C), while existing approaches support only 1-bit operations and exhibit poor thermal robustness. To address this, we propose a thermally robust multi-bit in-memory computing architecture. Our key innovation is a 2FeFET-1T memory cell enabling high-precision weight mapping and multi-bit multiply-accumulate (MAC) operations in the subthreshold regime. Integrated within a crossbar array and evaluated via NeuroSim simulation on VGG-8/CIFAR-10, the design achieves 91.31% classification accuracy—1.86% higher than the 1-bit baseline—and delivers 48.03 TOPS/W system energy efficiency, rivaling designs fabricated in more advanced technology nodes. This work represents the first demonstration of simultaneous multi-bit computation capability and strong temperature adaptability in subthreshold FeFET arrays, establishing a new paradigm for ultra-low-power edge AI acceleration.

Technology Category

• Machine Learning: Hardware-aware ML
• Cognitive Modeling & Cognitive Systems: (Computational) Cognitive Architectures
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Compute-in-memory (CiM) emerges as a promising solution to solve hardware challenges in artificial intelligence (AI) and the Internet of Things (IoT), particularly addressing the"memory wall"issue. By utilizing nonvolatile memory (NVM) devices in a crossbar structure, CiM efficiently accelerates multiply-accumulate (MAC) computations, the crucial operations in neural networks and other AI models. Among various NVM devices, Ferroelectric FET (FeFET) is particularly appealing for ultra-low-power CiM arrays due to its CMOS compatibility, voltage-driven write/read mechanisms and high ION/IOFF ratio. Moreover, subthreshold-operated FeFETs, which operate at scaling voltages in the subthreshold region, can further minimize the power consumption of CiM array. However, subthreshold-FeFETs are susceptible to temperature drift, resulting in computation accuracy degradation. Existing solutions exhibit weak temperature resilience at larger array size and only support 1-bit. In this paper, we propose TReCiM, an ultra-low-power temperature-resilient multibit 2FeFET-1T CiM design that reliably performs MAC operations in the subthreshold-FeFET region with temperature ranging from 0 to 85 degrees Celcius at scale. We benchmark our design using NeuroSim framework in the context of VGG-8 neural network architecture running the CIFAR-10 dataset. Benchmarking results suggest that when considering temperature drift impact, our proposed TReCiM array achieves 91.31% accuracy, with 1.86% accuracy improvement compared to existing 1-bit 2T-1FeFET CiM array. Furthermore, our proposed design achieves 48.03 TOPS/W energy efficiency at system level, comparable to existing designs with smaller technology feature sizes.

Problem

• Subthreshold FeFETs
• Temperature Variability
• Computational Precision

Innovation

• Temperature-resilient Computing-in-Memory
• Multi-bit 2FeFET-1T Design
• Energy Efficiency