top of page
Dan Meyer

Mass Communication-AT&T 5G Hitches Early Ride on IBM’s Satellite Cloud-B-AIM PICK SELECTS


AT&T is set to be one of the first passengers on IBM’s Satellite Cloud platform, with the telecom operator using that ride to bolster the reach and depth of its 5G network into enterprise markets. The move is also the latest by AT&T in tapping edge platforms from cloud providers to extend that network.

Jason McGee, VP and CTO for IBM Cloud Platform, explained that it was combining the vendor’s Cloud Satellite platform with AT&T’s 5G and edge work, with a focus on the carrier’s multi-access edge computing (MEC) efforts. The two will take this to market as an integrated offering that can be “dropped into customer environments.” Customers will then be able to use a single dashboard to manage services across multiple clouds and edge devices.

“That would include the kind of compute infrastructure to run services, to run applications, and the 5G infrastructure to give them connectivity,” McGee said. “And then satellite would provide a single pane of glass to manage that, and would give you the catalog of cloud services that you can then run in that environment. So you can deploy applications on those environments and manage them all from the cloud.”

That abstracted management is based on Red Hat’s Kubernetes-focused OpenShift platform that allows IBM to be flexible as to which infrastructure it can run the Cloud Satellite platform on.

“You obviously need some infrastructure to run the applications and services, but that infrastructure could be existing infrastructure that a customer already has like a set of (Red Hat Enterprise Linux virtual machines), or physical servers, or a VMware environment,” McGee said.

The AT&T model is a more integrated platform that IBM is also working on that is targeted at customers that might want a turnkey deployment model. This could include various forms of racks and servers that can support a Cloud Satellite location.

“So IBM would bring some of the infrastructure hardware, AT&T would bring that 5G infrastructure and we drop that configuration into one of these environments, and we use it to run,” McGee said.

And of course it would not be IBM if its Watson artificial intelligence (AI) platform was not involved, so Watson will be the source of insights in near real time from data collected at the edge.

Cloud Satellite Gains Altitude

IBM provided a peak into its Cloud Satellite service in May. It allows users to run their IBM Cloud service on-premises or in edge locations managed through a single pane of glass in the public cloud.

The platform uses IBM’s Cloud as a base and includes IBM Cloud Observability in order to provide a centralized view of running clusters and the services running in those clusters. Running on top of that base is a control plane that is designed to abstract away the operational details into a single management layer.

That set up can then be used to run and manage applications running in on-premises, hybrid edge, or infrastructure-as-a-service (IaaS) environments. Users can then use the systems to set up Cloud Satellite hosts to run in those locations. These hosts serve as a pool of resources designated to run the cloud services and applications and make up a satellite location. From there, these satellite locations can use what IBM calls a satellite link to link back from the cloud edge location to the control plane. The satellite link is not a real satellite link, but it does allow a user to manage networking requirements for the applications like firewalls and auditing.

On top of all of this is IBM’s managed services. The main platform will have the applications running on Red Hat’s OpenShift clusters that are managed by the control plane.

McGee noted that there had not been a lot of changes to the Cloud Satellite platform since the preview. He did explain that it was still in beta and that IBM was working toward a general availability release “early next year.”

“We continue to expand the set of services that we are able to run on Satellite, we are doing work around the Cloud Paks and our data portfolio Satellite,” McGee said. “AT&T is an example of what we’ll bring with some of our IoT and edge application management capabilities.”

AT&T and IBM Continue the Edge Focus

The combination also builds on AT&T’s ongoing work toward extending the reach of its 5G network. This includes a multi-year deal it struck with IBM more than a year ago that had AT&T Business become IBM’s primary provider of SDN and IBM helping AT&T improve and migrate its business applications to IBM Cloud.

AT&T more recently began deploying a 5G and MEC network at IBM’s lab in Yorktown Heights, New York, to develop and showcase new capabilities for enterprises seeking a private, on-premises cellular service. That service is also linked to IBM Cloud so that services and workloads running in that environment can coordinate with those running on the MEC platform at the facility.

While that deal is extensive, AT&T also has been working closely with Microsoft on a similar multi-year arrangement. That agreement, which was announced last year just days after the IBM deal, called for Microsoft to be the “preferred cloud provider” for non-network applications as part of the carrier’s broader push toward the cloud.

AT&T also has a deal with Hewlett Packard Enterprise (HPE) that will combine AT&T MEC services and HPE’s converged edge systems.

Those efforts are also running in tandem with similar moves by AT&T rival Verizon, which is working with both Microsoft and Amazon Web Services (AWS) on extending the reach of its 5G network.

McGee said that IBM’s work with Cloud Satellite has more flexibility than what some of the other vendors in the market are offering.

“We see it as an opportunity to uncover a bigger part of that spectrum and to push to the left and push toward those edge scenarios,” he explained. “So we designed an architecture that has flexibility at the infrastructure level so we can more quickly address these different variations, versus if we had picked an only integrated hardware approach then you would have had to have different variations of hardware and that takes time and there’s complexity there.”

Watch this video:https://www.youtube.com/watch?v=Z5vxRC8dMvs

Post: Blog2_Post
bottom of page