FAT: An In-Memory Accelerator With Fast Addition for Ternary Weight Neural Networks
- 17 June 2022
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
- Vol. 42 (3) , 781-794
Abstract
Convolutional neural networks (CNNs) demonstrate excellent performance in various applications but have high computational complexity. Quantization is applied to reduce the latency and storage cost of CNNs. Among the quantization methods, binary and ternary weight networks (BWNs and TWNs) have a unique advantage over 8 and 4-bit quantization. They replace the multiplication operations in CNNs with additions, which are favored on in-memory-computing (IMC) devices. IMC acceleration for BWNs has been widely studied. However, though TWNs have higher accuracy and better sparsity than BWNs, IMC acceleration for TWNs has limited research. TWNs on the existing IMC devices are inefficient because the sparsity is not well utilized, and the addition operation is not efficient. In this article, we propose FAT as a novel IMC accelerator for TWNs. First, we propose a sparse addition control unit, which utilizes the sparsity of TWNs to skip the null operations on zero weights. Second, we propose a fast addition scheme based on the memory sense amplifier (SA) to avoid the time overhead of both carry propagation and writing back the carry to memory cells. Third, we further propose a combined-stationary data mapping to reduce the data movement of activations and weights and increase the parallelism across memory columns. Simulation results show that for addition operations at the SA level, FAT achieves $2.00\times $ speedup, $1.22\times $ power efficiency, and $1.22\times $ area efficiency compared with a state-of-the-art IMC accelerator ParaPIM. FAT achieves $10.02\times $ speedup and $12.19\times $ energy efficiency compared with ParaPIM on networks with 80% average sparsity.
Keywords
All Related Versions
Funding Information
- Ministry of Education, Singapore, through Academic Research Fund Tier 2 (MOE2019-T2-1-071)
- Tier 1 (MOE2019-T1-001-072)
- Nanyang Technological University, Singapore, through NAP (M4082282)
- SUG (M4082087)
This publication has 53 references indexed in Scilit:
- Computing-in-memory with spintronics2018
- Fully parallel RRAM synaptic array for implementing binary neural network with (+1, −1) weights and (+1, 0) neurons2018
- Computing in Memory With Spin-Transfer Torque Magnetic RAMIEEE Transactions on Very Large Scale Integration (VLSI) Systems, 2017
- Emerging STT-MRAM circuit and architecture co-design in 45nm technology2017
- Deep Learning with Low Precision by Half-Wave Gaussian Quantization2017
- SCNNACM SIGARCH Computer Architecture News, 2017
- XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks2016
- EyerissACM SIGARCH Computer Architecture News, 2016
- Deep Residual Learning for Image Recognition2016
- FreePDK v2.0: Transitioning VLSI education towards nanometer variation-aware designs2009