An Evaluation of Edge TPU Accelerators for Convolutional Neural Networks

Kiran Seshadri, Berkin Akin, James Laudon, Ravi Narayanaswami, Amir Yazdanbakhsh

Abstract: Edge TPUs are a domain of accelerators for low-power, edge devices and are widely used in various Google products such as Coral and Pixel devices. In this paper, we first discuss the major microarchitectural details of Edge TPUs. Then, we extensively evaluate three classes of Edge TPUs, covering different computing ecosystems, across 423K unique convolutional neural networks. Building upon this extensive study, we discuss critical and interpretable microarchitectural insights about the studied classes of Edge TPUs. Mainly, we discuss how Edge TPU accelerators perform across convolutional neural networks with different structures. Finally, we present a learned machine learning model with high accuracy to estimate the major performance metrics of accelerators. These learned models enable significantly faster (in the order of milliseconds) evaluations of accelerators as an alternative to time-consuming cycle-accurate simulators and establish an exciting opportunity for rapid hardware/software co-design.