Abstract:The rapid development of edge artificial intelligence (AI) raises high requirements for data-intensive neural network (NN) computing and storage of edge devices, under a limited chip footprint and energy supply source. As a promising approach for energy-efficient processing, computing-in-memory (CiM) has been widely explored in recent efforts to mitigate the data transmission bottleneck. However, CiM with small on-chip memory capacity results in expensive data reloads, limiting its deployment in large-scale NN applications. Moreover, the increased leakage under advanced CMOS scaling lowers the energy efficiency. In this work, device-circuit synergy based on the indium-gallium-zinc-oxide (IGZO) thin-film transistor (TFT) is adopted to address these challenges. First, 4-transistor-1-capacitor (4T1C) IGZO eDRAM CiM is proposed with higher density than SRAM-based CiM and enhanced data retention by both lower device leakage and a differential cell structure. Second, exploiting the back-end-of-line (BEOL) compatibility and vertical integration of emerging channel-all-around (CAA) IGZO devices, 3D eDRAM CiM is proposed, which paves the way for IGZO-based CiM with ultra-high density. Circuit techniques including time-interleaved computing and differential refresh are proposed to guarantee accuracy under large-capacity 3D CiM. As a proof of concept, a $128$ $\times$ $32$ CiM array is fabricated under a foundry low-temperature poly-crystalline and oxide (LTPO) technology, demonstrating high computing linearity and long data retention. Benchmarks on scaled 45nm IGZO technology show energy efficiency of 686 TOPS/W for array only, and 138 TOPS/W while considering peripheral overheads.

A Computing-in-Memory Cell Design Based on LTPO Hybrid Thin Film Transistor Integration

A Low-Power In-Memory Multiplication and Accumulation Array with Modified Radix-4 Input and Canonical Signed Digit Weights

Low Power and Temperature-Resilient Compute-In-Memory Based on Subthreshold-FeFET

A Hybrid Computing-In-Memory Architecture by Monolithic 3D Integration of BEOL CNT/IGZO-based CFET Logic and Analog RRAM

Low-Power and Scalable BEOL-Compatible IGZO TFT eDRAM-Based Charge-Domain Computing

Low-Power and Scalable Retention-Enhanced IGZO TFT Edram-Based Charge-Domain Computing

Almost-Nonvolatile IGZO-TFT-Based Near-Sensor In-Memory Computing.

Computing-in-memory with thin-filmtransistors: challenges and opportunities

High-density Energy-Efficient Charge-Domain Computing Based on CAA-IGZO TFT with BEOL-compatible 3D Integration

One-Time-Programmable Memory in LTPS TFT Technology with Metal-Induced Lateral Crystallization

Computing-in-Memory with Thin-Film-Transistors: Challenges and Opportunities

Aging-Aware LTPO DTCO for Large-Scale Integrated Circuit-Driven Flexible Intelligent Sensing System

Counteractive Coupling IGZO/CNT Hybrid 2T0C DRAM Accelerating RRAM-based Computing-In-Memory Via Monolithic 3D Integration for Edge AI

LUNA-CIM: Lookup Table based Programmable Neural Processing in Memory

Monolithic 3-D Integration of Counteractive Coupling IGZO/CNT Hybrid 2T0C DRAM and Analog RRAM-Based Computing-In-Memory

Monolithic 3D Integration of Logic, Memory and Computing-In-Memory for One-Shot Learning

15.3 A 351TOPS/W and 372.4GOPS Compute-in-Memory SRAM Macro in 7nm FinFET CMOS for Machine-Learning Applications

Hardware Efficient Reconfigurable Logic-in-Memory Circuit Based Neural Network Computing.

TT@CIM: A Tensor-Train In-Memory-Computing Processor Using Bit-Level-Sparsity Optimization and Variable Precision Quantization

Fully Integrated 3-D Stackable CNTFET/RRAM 1T1R Array As BEOL Buffer Macro for Monolithic 3-D Integration with Analog RRAM-Based Computing-in-Memory