Litcius/Paper detail

A 28nm 276.55TFLOPS/W Sparse Deep-Neural-Network Training Processor with Implicit Redundancy Speculation and Batch Normalization Reformulation

Yang Wang, Yubin Qin, Dazheng Deng, Jingchuan Wei, Tianbao Chen, Xinhan Lin, Leibo Liu, Shaojun Wei, Shouyi Yin

202122 citationsDOI

Abstract

A dynamic weight pruning (DWP) explored processor, named Trainer, is proposed for energy-efficient deep-neural-network (DNN) training on edge-device. It has three key features: 1) A implicit redundancy speculation unit (IRSU) improves 1.46× throughput. 2) A dataflow, allowing a reuse-adaptive dynamic compression and PE regrouping, increases 1.52× utilization. 3) A data-retrieval eliminated batch-normalization (BN) unit (REBU) saves 37.1% of energy. Trainer achieves a peak energy efficiency of 276.55TFLOPS/W. It reduces 2.23× training energy and offers a 1.76× training speedup compared with the state-of-the-art sparse DNN training processor.

Topics & Concepts

Computer scienceNormalization (sociology)SpeedupParallel computingArtificial neural networkRedundancy (engineering)Artificial intelligenceComputer engineeringOperating systemSociologyAnthropologyAdvanced Neural Network ApplicationsAdvanced Memory and Neural ComputingFerroelectric and Negative Capacitance Devices
A 28nm 276.55TFLOPS/W Sparse Deep-Neural-Network Training Processor with Implicit Redundancy Speculation and Batch Normalization Reformulation | Litcius