The Role of Photonics for AI/ML at the Edge: What, Why, and How?


Hitesh Ballani, Microsoft Research Ltd., UK
Thomas Van Vaerenbergh, Hewlett Packard Ent., Belgium
S. J. Ben Yoo, Univ. of California Davis, USA


Edge Computing has become extremely important due to the rapid increases of data volume, low latency, privacy/security needs, 5G/6G. The ambition is to stop the need to rely on the cloud all the time, and process information at the location where it makes sense, which in some cases means where the data is generated.  To process this massive amount of data, advanced techniques from the AI/ML community are required, whereas photonics has a crucial role to play to increase throughput and reduce overall latency and energy-consumption. Examples of application domains with AI/ML opportunities in an Edge Computing context are ubiquitous: autonomous Vehicles, (Mobile) Medical Facilities, high frequency trading, (industrial) IoT based manufacturing and farming, 5G smart cell edge data center, AR/VR…

Our panellists will discuss how recent advances in photonics are critical to address some of the challenges that appear in this diverse set of workloads. 


To be determined.