Weak-MAC: Arithmetic Relaxation for Dynamic Energy-Accuracy Scaling in ConvNets
- 1 May 2018
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
Abstract
This work introduces Weak-MAC (WeMAC), a precision scaling strategy for fixed point Multiply&Accumulate operations. By leveraging an algorithmic relaxation of multiplication, it improves the efficiency of deep convolutional neural networks. WeMAC can be run on a software-programmable 8-bit MAC unit and is retraining free, namely, it makes ConvNet achieving reasonable inference accuracy even without retraining. Simulation results conducted on three ConvNets trained over well recognized data-set (SVHN, CIFAR-10 and CIFAR-100) demonstrate WeMAC enables new Pareto points in the energy-accuracy tradeoff of HW accelerators. With 25% more energy efficiency w.r.t. full precision, WeMAC represents a middle way between the low energy consumption of 8-bit arithmetic and the high accuracy of 16-bit arithmetic.Keywords
This publication has 8 references indexed in Scilit:
- 14.1 A 2.9TOPS/W deep convolutional neural network SoC in FD-SOI 28nm for intelligent embedded systems2017
- Understanding the impact of precision quantization on the accuracy and energy of neural networks2017
- DVAFS: Trading computational accuracy for energy through dynamic-voltage-accuracy-frequency-scaling2017
- 14.5 Envision: A 0.26-to-10TOPS/W subword-parallel dynamic-voltage-accuracy-frequency-scalable Convolutional Neural Network processor in 28nm FDSOI2017
- Going deeper with convolutions2015
- Deep learningNature, 2015
- Speech recognition with deep recurrent neural networks2013
- Symbolic synthesis of clock-gating logic for power optimization of control-oriented synchronous networks1997