Telcos May Not Even Try and Enter the General Purpose Computing as a Service Business at the Edge

It has to be said: in choosing to supply AWS with edge hosting facilities, a few tier-one telcos, likely to be followed by others, are making a considered bet that edge computing as a service is likely to be lead by, if not dominated, by the same providers in the ecosystem that dominate computing as a service. 

As they have found in other areas, winning a fight with Google, Facebook, Amazon and others in the application space is unlikely. That rational belief appears to condition strategies in the emerging edge computing space as well.

The AWS deal seems to signal belief that a general role as edge computing supplier will be lead by the hyperscalers. Hosting (the real estate role), on the other hand, might work. There are trade-offs. The highest-margin role likely will remain with the hyperscalers.

But a significant role in the hosting role would be a win for most telcos, who have generally not been able to carve out similar roles in the existing data center business.

It likely is too late for telcos to replicate the hyperscaler role in “as a service” computing, at the edge or elsewhere.

But edge data center hosting gives them another chance to carve out a role in the ecosystem. And the AWS partnership balances risk and reward, even if it signals belief in a smaller potential role in actual edge computing.

Telcos have not been hugely successful, outside of mobility or video entertainment, in creating big new businesses and revenue streams. Edge computing seems a promising area.

But many could recall that the data center business also was seen as a logical area of new revenue generation that meshed with the existing core competency of connections and data transport.

So recent deals between Amazon Web Services and a handful of tier-one service providers are instructive.

Basically, AWS Wavelengths creates AWS edge computing nodes at the edge of the telco network. So telcos act as providers of hosting (racks, power, security, cooling). 

In seeking that role in edge computing, the telco partners avoid the heavy capex required to emulate what the hyperscalers can provide their customers, instead choosing the simpler hosting role.

What they may be hoping is that the AWS moves lead to similar deals with many of the other hyperscale computing as a service providers, creating a data center hosting role some telcos tried and abandoned earlier. 

While other roles are not foreclosed, the AWS partnerships suggest that executives do not believe they are in position to invest in--or win--the battle for computing as a service. As many discovered earlier, the data center business has generally not been an area where telcos brought  significant advantages. 

On the other hand, perhaps many are betting that an early lead can be gained in the “edge facilities” part of the data center business, before potential rivals can scale their efforts. Of course, the hyperscale computing as a service suppliers are at the top of the list of potential leaders of the coming edge computing business. 

So the optimistic view might be that although not in position to lead edge-based computing as a service, telcos might secure a meaningful role in the edge data center hosting business, which requires distributed smallish data center locations.

Telcos of course have long considered former central offices or switching centers to be ideal real estate, in that regard. In metro areas where most of the edge computing demand will develop, central offices sit at the center of access networks running a few miles or so from end user locations. 

At least some mobile switching offices also are viable candidates for edge facilities as well.

Other roles are not foreclosed by the telco deals with AWS. It is conceivable that some vertical market services might develop where a few telcos are significant providers of the applications or capabilities. Vehicle communications and computing are logical candidates, for example. 

Still, the AWS deals are stark reminders that the edge computing ecosystem is, at the moment, most favorable for telcos as suppliers of rack space and communications. Most observers would probably agree with that assessment. 

That role also is arguably a capital-efficient and low-risk way  to enter the market. Other roles are not foreclosed. But perhaps few observers really believe the long-term telco opportunity is greatest anywhere but in the “pipeline” and “real estate” areas.