PCNN: Pattern-based Fine-Grained Regular Pruning towards Optimizing CNN Accelerators
Not on the map. Why?
Abstract (arXiv)
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.
Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). First posted on arXiv 11 Feb 2020.
Builds on
No reference list yet.
Built on by
Citing papers per year: 2021: 4; 2022: 4; 2023: 1.
9 papers.
- 2021Learning N:M Fine-grained Structured Sparse Neural Networks From Scratchcited by 177
- 2021SPViT: Enabling Faster Vision Transformers via Soft Token Pruningcited by 165
- 2022Enable Deep Learning on Mobile Devices: Methods, Systems, and Applicationscited by 32
- 2022Training Recipe for N:M Structured Sparsity with Decaying Pruning Maskcited by 10
- 2022Complexity-Driven CNN Compression for Resource-constrained Edge AIcited by 8
- 2023FPGA Resource-aware Structured Pruning for Real-Time Neural Networkscited by 7
- 2021Boosting the Convergence of Reinforcement Learning-based Auto-pruning Using Historical Datacited by 1
- 2021Neural Network Pruning Through Constrained Reinforcement Learningcited by 1
- 2022ClassPruning: Speed Up Image Restoration Networks by Dynamic N:M Pruningcited by 0
- 2023FPGA Resource-aware Structured Pruning for Real-Time Neural Networks9 Aug 2023
- 2022ClassPruning: Speed Up Image Restoration Networks by Dynamic N:M Pruning10 Nov 2022
- 2022Training Recipe for N:M Structured Sparsity with Decaying Pruning Mask15 Sep 2022
- 2022Complexity-Driven CNN Compression for Resource-constrained Edge AI26 Aug 2022
- 2022Enable Deep Learning on Mobile Devices: Methods, Systems, and Applications25 Apr 2022
- 2021SPViT: Enabling Faster Vision Transformers via Soft Token Pruning27 Dec 2021
- 2021Neural Network Pruning Through Constrained Reinforcement Learning16 Oct 2021
- 2021Boosting the Convergence of Reinforcement Learning-based Auto-pruning Using Historical Data16 Jul 2021
- 2021Learning N:M Fine-grained Structured Sparse Neural Networks From Scratch8 Feb 2021