AT&T's announced partnership with Microsoft last month was the telco's coming out party for its "public cloud first" strategy, which includes moving most of its non-network workloads to the public cloud by 2024.
Other key elements of the Microsoft deal, which is reportedly worth billions, include edge computing and a go-to-market strategy for the two large companies that share some of the same customers, according to Chris Rice, senior VP of network cloud and infrastructure at AT&T.
The multi-year alliance includes Microsoft being named as AT&T's preferred cloud provider for non-network applications as part of the telco's broader "public cloud first" strategy.
While AT&T is on track for virtualizing 75% of its network by next year, Rice said the deal with Microsoft included AT&T moving its non-network workloads, such as IT workloads and systems and HR functions, to the public cloud.
"Those virtualized network functions will remain on our private cloud, what we call Network Cloud," Rice said. "But there's a whole bunch of systems that are actually used to provision, to support, to do service management, to do fault management; all the other things that you do as part of delivering a service, all of those non-network workloads. Those will be the ones that we're starting to move to the public cloud.
"Then there are other things that are totally outside of network. There are other functions, kind of classic IT workloads, which will also have to be moved to the public cloud as well."
During AT&T's earnings call last month, company executives stressed that the cloud deal with Microsoft would help AT&T continue on the cost efficiency path that started with virtualization five years ago. Rice explained the cost efficiencies that the public cloud brings to AT&T as a "P × Q" component.
"There's a price and then there's a quantity," Rice explained. "So, think of, like, a unit, and a unit could be a unit of compute storage of network, and then there's a price for that unit, and then there's a quantity required to go run this function. So, you might need it, this Q quantity, for geographic reasons, for scaling reasons, for other reasons, and you basically build that Q quantity by whatever the busy hour is or the busy need is."
The Q quantity needs to go up during the launch of a new iPhone or a specific holiday, such as Mother's Day.
"The public cloud gives us the flexibility to move the Q part with that need," Rice said. "Whereas, if we were doing it on prem, we'd only really be able to handle that surge capacity by buying more functionality, by buying more of that quantity, and then leaving it in our data center until we need it again.
"So, in a way, that part of it is somewhat inefficient. Even though we run it efficiently for when we need it, we can't run it as efficiently because we don't have the ability to just scale up and scale down over shorter time periods. If you look at the fact that even if the price is a little bit higher during those shorter time periods, the fact that your average is much lower for the Q more than makes up for that price difference. So, that's one of the reasons why moving some of these systems and these IT functions to the public cloud makes sense for someone like AT&T who has variability in some of the workloads."
Rice said the IT functions and systems have run in AT&T's enterprise or IT cloud, but now they'll start to migrate to Microsoft's public cloud over the next four or five years.
What the public cloud means for the edge
In order to achieve the low latency—such as 20-millisecond round-trip latency— needed for new services and applications, the network has to be closer to the user. Rice said the edge is more about proximity and not connectivity.
AT&T has wireless and wireline connections to its customers. The powering, space and cooling are already in place in its network, and it has the field technicians that are needed to service those central offices or other network locations.
"Now, contrast that to the way the public cloud is built, which is primarily kind of like large, big data centers," Rice said. "So, they (cloud providers) have very big sites versus kind of the way we've done it, which is we have big sites but we also have many distributed sites. They also spend a fair amount of money and time on the user interface, the GUI. They spend money on the ecosystem supply and the marketplace.
"If you could marry the best of both worlds like that together, you have a way of going to the edge that we think makes sense. So part of the relationship that we'll have with them will be this joint way that we'll move to the edge together."
Customer and revenue opportunities from the deal with Microsoft
When it comes to developing business services and applications for enterprise customers, Microsoft and AT&T are creating a symbiotic relationship. AT&T can provide the go-to-market pipeline for Microsoft's cloud-based services and applications while the cloud enables new business use cases and revenues for AT&T.
"These are two very big companies with kind of a lot of different vectors associated with them," Rice said. "The reality is we're a very good network company, world-class arguably, and they're world-class as a software and cloud company. A lot of times putting solutions together that need all three of those turns into an integration effort for a lot of our customers.
"If we can do that right so we can go to market with what I will call pre-integrated solutions, it could be a lot better for our joint customers and since we have a very similar customer set that would makes sense."