Using Dataflow to Optimize Energy Efficiency of Deep Neural Network Accelerators

Reference Type:

Journal Article

Formatted Reference:

Published In:

IEEE Micro




Yu-Hsin Chen
Joel Emer
Vivienne Sze

Download Reference:

Search for the Publication In:


The authors demonstrate the key role dataflows play in the optimization of energy efficiency for deep neural network (DNN) accelerators. By introducing a systematic approach to analyze the problem and a new dataflow, called Row-Stationary, which is up to 2.5 times more energy efficient than existing dataflows in processing a state-of-the-art DNN, this work provides guidelines for future DNN accelerator designs.