A Layer-wised Mixed-Precision CIM Accelerator with Bit-level Sparsity-aware ADCs for NAS-Optimized CNNs
Open Access
- 20 January 2025
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 720-726
Abstract
Exploring multiple precisions as well as sparsities for a computingin-memory (CIM) based convolutional accelerators is challenging. To further improve energy efficiency with minimal accuracy loss, this paper develops a neural architecture search (NAS) method to identify precision for each layer of the CNN and further leverages bit-level sparsity. The results indicate that following this approach, ResNet-18 and VGG-16 not only maintain their accuracy but also implement layer-wised mixed-precision effectively. Furthermore, there is a substantial enhancement in the bit-level sparsity of weights within each layer, with an average bit-level sparsity exceeding 90% per bit, thus providing broader possibilities for hardware-level sparsity optimization. In terms of hardware design, a mixed-precision (2/4/8-bit) readout circuit as well as a bit-level sparsity-aware Analog-to-Digital Converter (ADC) are both proposed to reduce system power consumption. Based on bit-level sparsity mixed-precision CNNs benchmarks, post-layout simulation results in 28nm reveal that the proposed accelerator achieves up to 245.72 TOPS/W energy efficiency, which shows about 2.52 -- 6.57× improvement compared to the state-of-the-art SRAM-based CIM accelerators.Keywords
This publication has 9 references indexed in Scilit:
- LAMPS: A Layer-wised Mixed-Precision-and-Sparsity Accelerator for NAS-Optimized CNNs on FPGA2024
- IMPACT: A 1-to-4b 813-TOPS/W 22-nm FD-SOI Compute-in-Memory CNN Accelerator Featuring a 4.2-POPS/W 146-TOPS/mm2 CIM-SRAM With Multi-Bit Analog Batch-NormalizationIEEE Journal of Solid-State Circuits, 2023
- An Energy-Efficient Mixed-Bit CNN Accelerator With Column Parallel Readout for ReRAM-Based In-Memory ComputingIEEE Journal on Emerging and Selected Topics in Circuits and Systems, 2022
- 8-b Precision 8-Mb ReRAM Compute-in-Memory Macro Using Direct-Current-Free Time-Domain Readout Scheme for AI Edge DevicesIEEE Journal of Solid-State Circuits, 2022
- A High Performance Multi-Bit-Width Booth Vector Systolic Accelerator for NAS Optimized Deep Learning Neural NetworksIEEE Transactions on Circuits and Systems I: Regular Papers, 2022
- A 65 nm 73 kb SRAM-Based Computing-In-Memory Macro With Dynamic-Sparsity ControllingIEEE Transactions on Circuits and Systems II: Express Briefs, 2022
- STICKER-IM: A 65 nm Computing-in-Memory NN Processor Using Block-Wise Sparsity Optimization and Inter/Intra-Macro Data ReuseIEEE Journal of Solid-State Circuits, 2022
- Z-PIM: A Sparsity-Aware Processing-in-Memory Architecture With Fully Variable Weight Bit-Precision for Energy-Efficient Deep Neural NetworksIEEE Journal of Solid-State Circuits, 2021
- A Survey of the Usages of Deep Learning for Natural Language ProcessingIEEE Transactions on Neural Networks and Learning Systems, 2020