An Energy-Efficient Low-Latency 3D-CNN Accelerator Leveraging Temporal Locality, Full Zero-Skipping, and Hierarchical Load Balance
Hosted in Virtual Platform
DescriptionIn this paper, an energy-efficient low-latency three-dimensional convolutional neural network (3D-CNN) accelerator is proposed. Temporal locality and small differential value dropout are used to increase the sparsity of activation. Furthermore, to fully utilize the sparsity of weight and activation, a full zero-skipping convolutional microarchitecture is proposed. A hierarchical load-balancing scheme is also introduced to improve resource utilization. With the proposed techniques, a 3D-CNN accelerator is designed in a 55-nm technology, achieving up to 9.89x speedup compared to the baseline implementation. Benchmarked with C3D, the proposed accelerator achieves an energy efficiency of 4.66 TOPS/W at 100 MHz and 1.08 V.