WHILE EDGE SERVICE PROVIDERS PONDER SINGLE SOURCE VS.BEST OF BREED, CLOUD PLAYERS ARE POISED TO WIN
In edge services, service providers are still exploring where enterprise market demand will come from. Que visions of autonomous fleets and AR-assisted technicians. There is also the question of what edge services model has the upper hand: The one-stop-shop or best-of-breed combination? But in the big picture, cloud players make the differences between provider edge models less relevant.
Today, commercial enterprise test beds for edge services are still static. Early examples include wireless manufacturing floors, automated retail outlets, and healthcare facility monitoring systems. In these cases, edge compute resources could be located onsite as easily as hosted in the network. Network capacity, compute, and storage roll up to one place, and are more stable and predictable.
Edge is partly about untethering high-performance applications. In a nation or region, that might shift workloads from two or three centralized cloud locations to maybe 100+ distributed edge sites. Connecting devices with network-hosted edge needs network access, aggregation to an edge compute resource, the edge resource itself, and upstream connectivity to other cloud sites. All those components together should meet buyers’ real-time business intent. When there are so many variables, getting an outcome that matches business intent gets very complicated, very quickly.
On the bright side, providers are expert at modeling resources to anticipate peak loads. Service providers build infrastructure for the school year and major holidays, and support major broadcast events and major device vendor software updates. Each element in the network-hosted edge will be robust, handling wide swings. If delivery companies have a shipping crunch before a major holiday, or first responders work a localized emergency, the infrastructure holds.
But robust components do not mean connecting various edge elements is the sum of its parts. That is where a single provider that owns all the pieces—wireless and wired access, edge resources, connectivity up- and downstream—can claim to model traffic more predictably to business intent.
In terms of who has the upper hand, the unified and best-of-breed edge models each seem to have potential. A full-service provider might charge a premium for edge services needing stricter end-to-end performance. The full-service provider might leverage its premium installed base to package competitively priced, lower-priority bronze-level and best-effort edge plans. On the other hand, a large, carrier-neutral edge data center player could attract major tenants, forcing wireless and broadband access providers to connect at the best-performance they can, or else risk losing customers.
But the major cloud players can make it all a moot point. Cloud competitors are creating platforms to stretch across all assets considered edge. They seem poised to dominate, mixing infrastructure and wholesale services, negotiating contracts, setting supplier requirements, measuring performance, and tiering retail subscriptions by cost and performance. Cloud competitors add their applications’ marketplaces and retail clients. For all the attention network providers put on edge strategies, it seems likely their edge services role may be largely limited to infrastructure conduits.
Straight Talk is a weekly briefing from the desk of the Chief Research Officer. To receive this newsletter by email, please click here.
Contributed by knowledge partner: OMDIA