WidePipe: High-Throughput Deep Learning Inference System on a Cluster of Neural Processing Units

Lixian Ma, En Shao, Yueyuan Zhou, Guangming Tan. WidePipe: High-Throughput Deep Learning Inference System on a Cluster of Neural Processing Units. In 39th IEEE International Conference on Computer Design, ICCD 2021, Storrs, CT, USA, October 24-27, 2021. pages 563-566, IEEE, 2021. [doi]

Abstract

Abstract is missing.