Efficient Reprogramming of Memristive Crossbars for DNNs: Weight Sorting and Bit Stucking
This addresses the endurance issue in memristive hardware for AI applications, but it is incremental as it builds on existing crossbar architectures.
The paper tackles the problem of limited endurance in memristor-based compute-in-memory crossbars for deep neural networks by introducing weight sorting and bit sticking techniques, achieving a 3.7x reduction in reprogramming for ResNet-50 and 21x for ViT-Base on ImageNet-1K with accuracy loss within 1%.
We introduce a novel approach to reduce the number of times required for reprogramming memristors on bit-sliced compute-in-memory crossbars for deep neural networks (DNNs). Our idea addresses the limited non-volatile memory endurance, which restrict the number of times they can be reprogrammed. To reduce reprogramming demands, we employ two techniques: (1) we organize weights into sorted sections to schedule reprogramming of similar crossbars, maximizing memristor state reuse, and (2) we reprogram only a fraction of randomly selected memristors in low-order columns, leveraging their bit-level distribution and recognizing their relatively small impact on model accuracy. We evaluate our approach for state-of-the-art models on the ImageNet-1K dataset. We demonstrate a substantial reduction in crossbar reprogramming by 3.7x for ResNet-50 and 21x for ViT-Base, while maintaining model accuracy within a 1% margin.