◎ Abstract 

There is much interest in embedding data analytics and artificial intelligence capabilities into sensor-rich platforms such as wearables, biomedical devices, autonomous vehicles, robots, and the Internet of Things (IoT) to provide them with decision-making capabilities. Such platforms need to implement machine learning algorithms under severe resource constraints on embedded batterypowered platforms. As a result, there have been on-going effort to realize energy- and latency- efficient information processing. In this talk, recent research trend in machine learning accelerator architecture will be discussed including algorithm, circuit, and architecture level innovations. In addition, both digital and analog hardware design will be discussed to accelerate recently-emerging Transformer network models.