Multiple-Deep Neural Network Accelerators for Next-Generation Artificial Intelligence Systems

Stylianos I. Venieris, Christos-Savvas Bouganis, Nicholas D. Lane

Abstract: The next generation of artificial intelligence (AI) systems will have multi-deep neural network (multi-DNN) workloads as their core. Large-scale deployment of AI services and integration across mobile devices require additional breakthroughs in the computer architecture front, with processors that can maintain high performance as the number of DNNs increase, giving rise to the topic of multi-DNN accelerator design.