Verizon applied edge compute technology to cut latency in half in a trial conducted on a live 5G network in Houston, Texas.

Using multi-access edge compute (MEC) equipment placed near the edge of the network at its 5G test bed, the operator said a facial recognition application was able to identify individuals twice as fast as when it ran on a network which processed information at a centralised data centre. Verizon did not provide a latency measurement from the test.

The trial follows Verizon’s move in 2018 to lay the foundation for MEC. In April 2018, it began the process of decoupling software and hardware on routers at the edge of its network. Part of its so-called Intelligent Edge Network initiative, Verizon executives at the time told Mobile World Live the change would allow it to scale its edge hardware and software separately, and deliver the ability to launch new services more quickly.

Edge benefits
Verizon has long talked up ultra-low latency as a key benefit of 5G: CEO Hans Vestberg (pictured) recently highlighted the capability as one of eight pillars of 5G which will enable new use cases in areas such as virtual reality, industrial automation, autonomous cars, healthcare and gaming.

But Verizon is looking to reap other gains beyond latency as it builds out its MEC network. It noted in a press release “an increase in reliability, energy efficiency, peak data rates and the ability to process more data through more connected devices are also benefits of introducing MEC technology.”

Rival AT&T has similarly been moving toward an MEC architecture, rolling out open source whitebox hardware at its cell sites in 2018 as part of its transition. It also opened an edge computing test bed in Silicon Valley to explore potential use cases, and partnered with The Linux Foundation on an open source project to create the software stack necessary to support edge computing.