Weight pruning is a powerful technique to realize model compression. We propose PCNN, a fine-grained regular 1D pruning method. A novel index format called Sparsity Pattern Mask (SPM) is presented to encode the sparsity in PCNN. Leveraging SPM with limited pruning patterns and non-zero sequences with equal length, PCNN can be efficiently employed in hardware. Evaluated on VGG-16 and ResNet-18, our PCNN achieves the compression rate up to 8.4X with only 0.2% accuracy loss. We also implement a pattern-aware architecture in 55nm process, achieving up to 9.0X speedup and 28.39 TOPS/W efficiency with only 3.1% on-chip memory overhead of indices.
@article{arxiv.2002.04997,
title = {PCNN: Pattern-based Fine-Grained Regular Pruning towards Optimizing CNN Accelerators},
author = {Zhanhong Tan and Jiebo Song and Xiaolong Ma and Sia-Huat Tan and Hongyang Chen and Yuanqing Miao and Yifu Wu and Shaokai Ye and Yanzhi Wang and Dehui Li and Kaisheng Ma},
journal= {arXiv preprint arXiv:2002.04997},
year = {2020}
}