Just as companies once began to hire cloud service providers to abstract away as much computing and storage as possible, they are now beginning to do the same with networks.

The enterprise is starting to look beyond ordinary routers and switches to a network-centered approach that interconnects applications, offices, employees and users.

“I think the cloud providers individually are abstracting away as much of the network as they possibly can,” said Caroline Chappell (pictured left), research director at Analysys Mason Ltd. “Cloud providers want to appear to developers just as some kind of plumbing. And it’s very easy now for enterprises to say, ‘OK, I want this service and I’m just going to go through their API and connect to it.’ And that’s why you get in a situation of multiple, multiple clouds.”

Chappell and Andrew Coward (pictured right), general manager of software-defined networking at IBM, spoke with John Furrier (@furrier), host of theCUBE, SiliconANGLE Media’s livestreaming studio, during IBM Think. They discussed why networking is more important than ever for IBM, the new hybrid cloud and edge computing reality, and the multicloud future. (* Disclosure below.)

[Editor’s note: The following content has been condensed for clarity.]

Why is networking now more important than ever for IBM? 

Coward: Networking is weaved into pretty much everything we touch — from Red Hat Linux, data analytics, machine learning tools, security, cloud services and so on. And the networking business is changing very radically at the moment. We’re going through a kind of massive shift, not just to the cloud, but the desegregation of networking products, which you would think of being very tight and integrated, are actually being separated into their constituent parts — distribution of applications and data across multiple clouds.

The other thing … is that there are like 15 billion network-capable devices out there with general computing capabilities. And all of those are expected to connect into the public or private cloud. And so how they connect is a critical concern to everything we do at IBM. 

Caroline, now the conversation is hyper-focused on cloud integration at-scale with the kind of features that enterprises really need. What’s your take on all this radical change?  

Chappell: The network itself is becoming disaggregated into hardware and software. It is really becoming a software application that runs on the cloud itself. And that means you can distribute the network in a very different way than you could do in the past. And what that’s really affecting is who can provide a network, how they can provide it, and what network services they can provide. And I think that is changing the decision points for operators [and] for enterprises. They’re being faced with a very big choice about who will provide their connectivity services? Will it be an SD-WAN vendor? Would it be a SaaS player that’s basically just operating out in the cloud. And if you look at the services themselves, there’s the opportunity for enterprises to build really kind of rich, bespoke connectivity on demand and in a way that they’ve never had before. 

What’s the disruption, and what’s it mean for the enterprise networks over the next couple of years going forward? 

Chappell: I think that the idea of the network as being something fixed, persistent with fixed persistent connections is changing. A lot of the new stuff that’s coming along with the IoT-driven stuff, a lot of the changes around the edge, and operational process automation will ask for on-demand connectivity. 

A lot of the applications themselves will run on the cloud, and not just on one cloud, but on many distributed clouds. So you’ve got to think about zero-trust security, because you are basically spinning up these connections on demand. So I think enterprises have got to deal with those data and security and all their best practices. We’ve got to shift to a much more dynamic, connectivity world.

Andrew, I want you to weigh in on the IBM impact, because what we just heard was application-driven — that’s DevOps, that’s programmability. Now you’ve got DevSecOps. All this is now the requirement. What’s the bet on the IBM side? 

So, where’s that data allowed to move to and from, and what are the boundary points for that? How is the reliability, resiliency of our system guaranteed across many disparate parts? All of those things end up being a very vertically integrated solution. But fundamentally we’ve got a very different way of being able to react dynamically to both the network, the application, and ultimately the end user.

Watch the complete video interview below, and be sure to check out more of SiliconANGLE’s and theCUBE’s coverage of IBM Think. (* Disclosure: TheCUBE is a paid media partner for IBM Think. Neither IBM, the sponsor for theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE