Technology pundits and media proclaimed that 2019 would be the year of 5G, then reasserted one year later that 2020 would be THE year. As we wrap up 2020, the same folks declare that 2021 is the year 5G will live up to its hype. 5G used to be the source of national pride. Remember the April 2019 fracas between U.S., China and South Korea MNOs fighting for the mantle of world's first commercial 5G rollout? Today, consumers appeared uninterested, unfazed and confused by the super-hyphenated qualified firsts — "world's-fastest-5G-mid-band-for-consumers-within-100-miles-of-any-U.S.-metro-area-with-more-than-1M-in-population-per-U.S.-2010-census."
Covid-19, which put major economies on pause or even rewind, nary put a dent on worldwide 5G rollouts. And yet, I'm expecting a repositioning away from pure 5G messaging as we enter 2021. Edge computing, which has been closely associated with 5G, has experienced a similar journey. Much of this is typical of any technology innovation lifecycle — Geoffrey Moore's chasms, bowling alleys, and more, or Gartner's hype cycle. So, what will 2021 bring for edge computing (associated with 5G and otherwise)? Before I provide AvidThink's perspectives, let's do a quick look back on what the industry achieved in 2020.
2020 edge computing highlights
In 2020, in conjunction with the buildout of mobile 5G networks and fiber assets, both wireless and wireline communications service providers (CSPs) continued to make investments in edge computing. From BT to Telstra, KT to KDDI, China Unicom to Telefónica, and Lumen (CenturyLink), wireline and wireless providers rolled out trial edge sites at varying scales. For instance, Lumen (CenturyLink) rolled out 100 bare-metal edge sites, with another 150 on their way. Lumen is also supplementing that rollout with VMware and Morpheus Data partnerships to provide support for hybrid edge and cloud management.
Other carriers rolled out edge computing in conjunction with hyperscale cloud providers like AWS. As of December 2020, AWS has rolled out its generally available (GA) Wavelength solution with its telco partners: Verizon (eight sites in the U.S. with two more by year-end), KDDI (one site in Japan), SKT (one site in Korea), Vodafone (UK and Germany - pre-GA). AWS has also rolled out its Local Zones edge service directly to end-users (three locations in the U.S. with 13 more in 2021). AT&T and Microsoft Azure have promised three sites in the U.S. but have yet to launch. Google Cloud has announced numerous worldwide telco partnerships under its GMEC umbrella with no launches as yet.
Hyperscalers did not limit their edge plays to telco partnerships, with AWS and Azure launching enterprise on-premises edge offering: AWS Outpost and Azure Private Edge Zones. Both can be orchestrated from the cloud but benefit from local data privacy and lower latencies due to their on-campus locations. In these deployments, carriers can play a partner role, using these on-premises edges to host and operate private enterprise networks based on private LTE or 5G. For instance, Telefónica Germany uses AWS to power private enterprise networks with Ericsson's 5GC mobile core. Carriers have embarked directly on their on-premises edge strategy, sometimes partnered with network equipment providers (NEPs) like Nokia, Ericsson and Cisco, or integrators and ISVs like IBM/Red Hat and sometimes on their own.
Not to be outdone by the messy collection of carriers, hyperscalers, and NEPs, a range of startups, data center operators, co-location providers, and even tower companies are muscling in on edge action. Datacenter company Switch is building out modular edge data centers at FedEx facilities in the U.S. Vapor IO and Digital Realty have partnered to build out edge sites (three launched in the U.S. so far.) Equinix acquired Packet, and it now powers their bare-metal offerings across multiple metro locations. American Tower launched six edge data center sites in the U.S. on ground space near existing power locations.
The cable MSOs have also jumped into the fray. Cox Communications (with Juniper Networks) funded Stackpath — a startup building out edge computing locations with integrated network security and caching services.
2021 edge computing outlook
The edge computing ecosystem is unlikely to consolidate in 2021. Everyone with a claim on the space will continue to push: wireline operators, MNOs, hyperscalers, software vendors, MSOs, tower companies, co-location providers, and more. I expect to see the bulk of action focused on the hyperscalers and telcos intersection.
AWS and its partners have aggressively rolled out Wavelength, and I expect continued rollouts of Wavelength and Local Zones in 2021. AT&T and Microsoft Azure will almost certainly launch their joint mobile edge offering in 2021, and Azure's other global telco partners will follow with joint launches.
Google continues to work on its edge initiative, focusing initially on an orchestration story with Anthos, now spinning a hybrid multi-cloud story, with a liberal expansion of its partner ecosystem. There have been no concrete launches as yet. Nevertheless, for a hyperscaler with extensive edge placements in the form of global caches used for YouTube and other services like Stadia gaming, we'd expect to see those sites repurposed in 2021 as edge clouds. Even CDNs like Limelight and Akamai will get into the edge skirmish.
The hyperscaler cloud battle will extend into the enterprise on-premises edge. AWS Outpost, Azure Private Edge Zones, and Google will push against platforms from HPE, IBM and Lenovo coupled with software from VMware, Red Hat, and a host of startups. The coopetition between ecosystem players will confuse enterprises trying to find a clear path forward, and I don't expect relief on this front in 2021. Some edge initiatives will founder and collapse, such as Ericsson's Edge Gravity did earlier this year, but more will arise to replace them, adding to the list of edge choices.
On the carrier front, we may see some traction on the GSMA Operator platform or the 5G Future Forum on its MEC initiatives. But I'm not holding my breath. Figuring out MEC use cases are hard enough without coordinating between multiple major global telcos.
Forging edge value chains in 2021
There are now generally available commercial offerings for the mobile edge, wireline edge, and on-premises enterprise edge. Therefore, I expect that 2021 will see multiple value chains forming as enterprises look to solve their business challenges. It'll take another year or two for the industry to figure out which value chains are stable and likely at the same time to figure out which edge use cases make sense. Running workloads at the edge comes with increased costs (compared to regional hyperscaler clouds) and a reduced set of available supporting services but provides improved data controls and lower latencies.
It's unclear whether hyperscale cloud providers will dominate the edge due to their robust developer relationships, distributed cloud platform, scale economies and large cash war chests. There's certainly a possibility that virtualization software vendors like VMware and IBM/Red Hat could succeed in abstracting away underlying public/private clouds, reducing hyperscaler strangleholds. In any case, carriers will almost certainly run a multi-cloud playbook, leveraging one hyperscaler against another to reduce their supplier power.
Carriers will also try to build their edge platforms, sometimes in conjunction with innovative ISVs. Still, I'm concerned about the carriers' ability to connect with and convince developers to cast their lot with them. Likewise, it's unlikely that carriers' experiences in deploying cloud-native network functions or VNFs, or open RAN functions will give them sufficient learning or economies of scale to counter hyperscaler experience in edge and clouds.
Further, there's currently no open-source edge platform from Linux Foundation, Eclipse Foundation, OpenInfra Foundation (OpenStack), or other open-source groups with sufficient momentum or scope to counter the hyperscalers.
What that means in 2021 for carriers and enterprises alike is that many edge offerings will involve the hyperscalers — and as Tess Uriza Holthe elegantly puts it: "When the elephants dance, the chickens must be careful."
Roy Chua is founder and principal at AvidThink, an independent research and advisory service formed in 2018 out of SDxCentral's research group. Prior to co-founding SDxCentral and running its research and product teams, Chua was a management consultant working with both Fortune 500 and startup technology companies on go-to-market and product consulting. As an early proponent of the software-defined infrastructure movement, Chua is a frequent speaker at technology events in the telco and cloud space and a regular contributor to major leading online publications. A graduate of UC Berkeley's electrical engineering and computer science program and MIT's Sloan School of Business, Chua has 20+ years of experience in telco and enterprise cloud computing, networking and security, including founding several Silicon Valley startups. He can be reached at [email protected]; follow him at @avidthink and @wireroy
Industry Voices are opinion columns written by outside contributors—often industry experts or analysts—who are invited to the conversation by FierceTelecom staff. They do not represent the opinions of FierceTelecom.