The Competitive Landscape of Edge Acceleration Chips for Deep Learning
IDC forecasts that the market of artificial intelligence (AI) accelerators for internet of things (IoT) edge and primary clients is about to reach 6.1 billion in shipment by 2020. This document describes the competitive landscape of edge accelerators for neural networks, including the stellar players, major competitions, powerful new entrants, leading IP licensers, young start-ups, and experimenters of alternative technologies. "Many businesses have come to realize that there are choices of acceleration technologies for AI," says Jessie Cai Danqing, senior research manager of big data, artificial intelligence, and content analytics practice, IDC Asia/Pacific. She adds, "Application-specific designs to lower power consumption, cost, and size could open up opportunities for vendors to challenge the leader position of Nvidia in the AI acceleration market, especially for edge workloads on the mass market of primary clients, surveillance cameras, drones, automobiles, and industry systems, including healthcare, commercial transportation, energy, and industrial."