A 40-nm MLC-RRAM Compute-in-Memory Macro With Sparsity Control, On-Chip Write-V erify, and Temperature-Independent ADC References Wantong Li ,X i a o y uS u n , Shanshi Huang

Abstract

Resistive random access memory (RRAM)-based compute-in-memory (CIM) has shown great potential for accel- erating deep neural network (DNN) inference. However, device characteristics, such as low-resistance values, susceptibility to drift, and single-level cells, may limit the capabilities of RRAM-based CIM. In addition, prior works generally used the off-chip write-verify scheme to tighten RRAM resistance distributions and used off-chip analog-to-digital converter (ADC) references for fine-tuning partial sum quantization. Although off-chip techniques are viable fo r testing purposes, they may be unsuitable for practical applications. In this work, we present an RRAM-CIM macro to accelerate DNN inference. The chip features: 1) multi-level cell (MLC) RRAM for improving compute performance and density; 2) sparsity-aware input control to leverage the high activation sparsity in DNN models; 3) on-chip write-verify to speed up initial weight programming and peri- odically refresh cells to compensate for resistance drift under stress; and 4) on-chip ADC reference generation that provides column-wise tunability and stability with varying temperatures to guarantee the CIFAR-10 accuracy of 85.8% at 120 ◦C. The design is fabricated in TSMC 40-nm process with embedded RRAM technology and achieves a macro-level peak performance of 97.8 GOPS/mm2 and 44.5 TOPS/W for multiply-and-accumulate (MAC) operations on VGG-8 net work with ternary weights.