9.2 A 28nm 12.1TOPS/W Dual-Mode CNN Processor Using Effective-Weight-Based Convolution and Error-Compensation-Based Prediction
Huiyu Mo, Wenping Zhu, Wenjing Hu, Guangbin Wang, Qiang Li, Ang Li, Shouyi Yin, Shaojun Wei, Leibo Liu
Abstract
To deploy convolutional neural networks (CNNs) on edge devices efficiently, most existing CNN processors were built on quantized CNNs to optimize the inference operations. However, three issues (Fig. 9.2.1) have not been well addressed: 1) Duplicate weights in each kernel after quantization yielding repetitive multiplications; 2) a huge number of unnecessary MACs caused by ReLU activation functions; 3) frequent off-chip memory access in residual blocks.
Topics & Concepts
Computer scienceQuantization (signal processing)Convolutional neural networkConvolution (computer science)Kernel (algebra)ResidualEdge deviceEnhanced Data Rates for GSM EvolutionParallel computingAlgorithmArtificial neural networkArtificial intelligenceMathematicsDiscrete mathematicsOperating systemCloud computingAdvanced Neural Network ApplicationsAdvanced Memory and Neural ComputingMachine Learning and ELM