Previous Webinars – Computing for AI is undergoing a major shift, from a cloud-centered paradigm to a more edge-centric one. Edge computing makes low power consumption of devices necessary. In addition, data centers that host cloud servers consume enormous amounts of electricity. Therefore, development of AI hardware that can reduce power consumption both for the cloud and at the edge is pressingly needed. This webinar will pinpoint which component of AI hardware is the most power hungry, and introduce hardware solutions that can effectively reduce its power consumption.