Featured Story
AT&T struggles to defend open cloudiness of Ericsson deal
More than a year into the Ericsson-led rollout, there is very little evidence AT&T's radio access network is as multivendor and virtualized as the telco makes out.
Once a major initiative from the cable company, Cox Edge has been folded into the carrier's private networks business – another signal that edge computing continues to struggle.
Cox Edge – the edge computing service launched by the cable company in 2021 – is no longer available from the company as a discrete business.
Cox representative Todd Smith said the company folded its Cox Edge brand and capabilities into its Cox Private Networks business group about a year ago. Cox Private Networks, launched in 2023, is part of Cox's Hospitality Network team and builds wireless networks for convention centers, hotels and other large venues.
Smith said Cox still owns the roughly 30 different computing sites, scattered across the US, that comprised Cox Edge. He said those sites are now part of the capabilities provided within Cox's private networks business.
However, Smith did not respond to follow-up questions about those edge computing capabilities. Cox no longer operates its Cox Edge website at www.coxedge.com, and the website for its Cox Private Networks business makes almost no mention of edge computing.
Cox Edge initially launched with content delivery network services, edge bare metal offerings, serverless connectivity, virtual machines, distributed databases and managed Kubernetes products. At first it spanned a dozen sites in cities like Seattle and Denver, and ultimately expanded to almost 30 cities around the US – with plans to eventually double that number. Ron Lev initially headed the business as its GM; he left Cox in 2023 and is now head of edge at NTT Data, according to his LinkedIn profile.
A retreat
Of course, Cox is not alone in retreating from the edge computing market.
"The mobile edge compute ... the adoption curve [is] a little slower than maybe we would like," explained a top Verizon executive in 2023.
Verizon kicked off a major edge computing initiative with Amazon Web Services (AWS) in 2019, but has struggled to generate demand for the offering. Verizon though still touts its private and public edge computing operations on its website. So does Lumen Technologies, which first announced its edge efforts in 2020.
"STL Partners has tracked companies that are winding down their network edge propositions, having already built capacity. This is the case for AT&T (which had deployed two sites) and Cox Communications (which had deployed 30 sites)," the analyst firm wrote late last year.
During the 2010s, a wide number of companies were planning to ramp up major investments into the edge computing industry. The reasoning for such deployments is relatively simple: Small, unmanned data centers in modest cities would be the only way to provide super low-latency services to residents in such locations. Otherwise, their Internet traffic would have to travel all the way to bigger data centers in Denver or Dallas, adding precious milliseconds to services like streaming virtual reality that basically need to be instantaneous.
The coronavirus pandemic, which started in 2020, put all those efforts on hold.
A return to the edge
Despite the failure of edge computing to catch fire – particularly among service providers like Cox – a number of analyst firms continue to predict that demand for edge computing sites will grow. AI is their primary reason.
For example, STL Partners this week predicted that the global addressable edge AI market is set to grow by a compound annual growth rate of 19% during the next six years, reaching $157 billion in revenue by 2030.
"The substantial value generated directly by AI or as a direct enabling force indicates that players which are specialising in accommodating AI applications at the edge stand to gain significant value," explained George Glanville, a research analyst at STL Partners and author of the forecast, in a release.
Others agree.
Marc Ganzi, the CEO of Digital Bridge, cited "strong AI-led demand for more data center power and capacity" for his company's recent transaction with DataBank. DataBank touts itself as "the largest edge data center operator in the US," with more than 65 data centers and 20 interconnection hubs in more than 27 markets across the country.
AI "inference" is the driving factor here. Inferencing in AI refers to the process where a trained AI model makes predictions or decisions based on new data. It's essentially AI's ability to apply learned knowledge to fresh inputs in real-time. Edge computing plays a critical role here because it brings inferencing closer to users. This lowers latency (meaning, instant AI responses) and can also reduce bandwidth costs and ensure privacy and security.
You May Also Like