上一条： PRArch: Pattern-Based Reconfigurable Architecture for Deep Neural Network Acceleration

下一条： Energy-efficient and quality-assured approximate computing framework using a co-training method