Everyone loves the cloud for its convenience and ability to host and hold just about anything you can think of, but there are places where it falls drastically short. Latency becomes more critical every year to network performance, with AI in full bloom and AR/VR making its steady way into broader use. As latency concerns grow, so does the proliferation of edge computing.
AI got a lot of mindshare time from CES 2024, where everyone was linking the technology from the latest phones to an AI pillow (I kid you not) and a separate AI mattress. Google and Samsung’s latest phones are adding onboard AI features, arguably a very sharp example of edge computing on the phone pairing with cloud computing to deliver a much better and more useful experience.
Edge computing simply puts the necessary processing resources as close as possible at the edge of the network to perform tasks and generate results that are dependent upon low latency. One example is the use of edge computing in self-driving cars. If you are driving down the road and the cameras and sensors in your car are being processed to provide warning information about crossing over into another lane or a vehicle in your blind spot before you change lanes, that’s a perfect example of edge computing taking place in real-time. Going to the cloud is not an option because you might not have microsecond level connectivity, and you need the results (car in a blind spot) to act on immediately, not a half second or a couple of seconds later.
Sticking with connected vehicles, another edge compute application is emerging. McKinsey predicts that there will be 1.8 billion connected vehicles in 2030 with each car generating 1 to 2 terabytes of raw data each day. Much of this data contains personally identifiable information about the drivers and passengers. Due to privacy concerns, a great amount of this data must be filtered at the edge before being uploaded to the cloud. That’s where I see an application for fiber fed edge compute nodes collocated with electric vehicle charging stations.
Another example of the two-way flow of data for edge computing is Tesla’s ongoing evolution of AI, with data collected from its in-house fleet of self-driving cars collected and fed into the company’s machine learning models to make better decisions. The new models are then rolled out from Tesla’s AI cloud, travel through fiber and end up updating the larger publicly owned fleet of vehicles around the world.
Some will point out that these examples rely on wireless connectivity, which is true. But in truth, most wireless networks are just fiber networks with wireless antennas on the ends Fiber may not be in the last foot, but it definitely is present in connecting that last foot to the rest of the world. With edge computing coming into play for everything from phones to cars to localized computing resources to reduce overall latency, fiber’s story as the key unsung role in enabling today’s information society continues.
Kevin leads the marketing efforts for Clearfield as Chief Marketing Officer. He joined the fiber company in 2016, leveraging his extensive experience in advanced communications technology, fiber optic systems, and business product marketing. Prior to joining Clearfield, he spent two decades serving in various senior marketing positions at ADTRAN. Before that, he spent a decade at telephone operating company BellSouth, now a part of AT&T, where he worked as the lead broadband product evaluations resource in the Science & Technology department.
Morgan currently serves on the Fiber Broadband Association Senior Council Committee and has also held various leadership positions at the Fiber Broadband Association, including Board of Directors Chair for 2015, 2019, and 2022. Morgan holds an Electrical Engineering degree from Auburn University and an MBA from the University of Alabama.