EPIM: Efficient Processing-In-Memory Accelerators based on Epitome
This addresses memory constraints in PIM accelerators for neural network deployment, offering a novel operator design that surpasses pruning methods, though it is incremental in improving hardware efficiency.
The paper tackles the challenge of limited on-chip memory in Processing-In-Memory (PIM) accelerators for large-scale neural networks by introducing EPIM, a method using lightweight epitome operators and hardware modifications, achieving 71.59% top-1 accuracy on ImageNet with 3-bit quantization and reducing crossbar areas by 30.65 times.
The utilization of large-scale neural networks on Processing-In-Memory (PIM) accelerators encounters challenges due to constrained on-chip memory capacity. To tackle this issue, current works explore model compression algorithms to reduce the size of Convolutional Neural Networks (CNNs). Most of these algorithms either aim to represent neural operators with reduced-size parameters (e.g., quantization) or search for the best combinations of neural operators (e.g., neural architecture search). Designing neural operators to align with PIM accelerators' specifications is an area that warrants further study. In this paper, we introduce the Epitome, a lightweight neural operator offering convolution-like functionality, to craft memory-efficient CNN operators for PIM accelerators (EPIM). On the software side, we evaluate epitomes' latency and energy on PIM accelerators and introduce a PIM-aware layer-wise design method to enhance their hardware efficiency. We apply epitome-aware quantization to further reduce the size of epitomes. On the hardware side, we modify the datapath of current PIM accelerators to accommodate epitomes and implement a feature map reuse technique to reduce computation cost. Experimental results reveal that our 3-bit quantized EPIM-ResNet50 attains 71.59% top-1 accuracy on ImageNet, reducing crossbar areas by 30.65 times. EPIM surpasses the state-of-the-art pruning methods on PIM.