IBM: ‘There are not many firms in the world that really understand the telco sector’

In part two of our interview with tech giant IBM, Stephen Rose, IBM’s global General Manager for TME and Distribution talks Open RAN, edge computing, and the future of the telecoms industry.

How do you see the current state of Open RAN in its present form and how large a part of the telecoms industry will it be in the next ten years?

It’s important to appreciate how far the discussion has actually come. The discussion on the notion of Open RAN and open interfaces has been on the cards for 20 years, it’s only in the last three to five years we’ve really seen it actually take shape, and it has taken shape in a couple of ways. Of course, we’ve become organised better around bodies and standards. The other element to that is that the OEMs have started to partake in that discussion much more. And the key for Open RAN is making sure that you can get the performance of these Open RAN networks at the equivalency of the dominant OEM players – Nokia and Ericsson and whoever else is out there right now.

The other element is to ensure that we understand how those elements move between vendors, as the as the disaggregation of the of the architecture happens. I think primarily the concern for the industry is getting the architectural start point established first and making sure that the disaggregated hardware and software even within a monolithic structure can work as it could do in a bare metal environment. That’s the basic table stakes. You’ve got to you’ve got to make sure you get that right.

As we then progressively move in the performance elements, companies like IBM can then play a role where we can then stand up these different environments in our labs, and we can test between different types of vendors and ensure that two things are happening. One is that the performance at the element level is happening, but then the performance across the entire stack is happening. I think that’s incredibly important, especially as we go towards high demand services.

So I think there could be a couple of steps that happen in the industry, it could be that we see Open RAN deployed concurrent to traditional architectures, or within disaggregated monolithic environments. And you’ll see that if the industry can’t quite get the performance that is looking for then it may be that you’ll see unique deployment of the technology at an earlier stage, and then progressively go towards more of a of an industry standardised way of running the technology for all different types of use cases and services.

I think it’s an exciting space, I think it does drive innovation. The initial imperative, around 30% TCO drop, was over ambitious simply because if you focus only on the money, you actually drive away the incentive in the innovation. But I think if we if we focus on the notion that additional players into the marketplace drive up the innovation, and we could then deploy the technology progressively over time then it’s got a real shot of being successful.

As we move towards 6G, it will become extremely standard. Where do I think it will be 10 years from now? I think it could be up to 25% market share. What it really takes is one of the OEMs to move forward and make a not only a definitive step verbally, but also in real working practice. When one of the OEMs puts big money into R&D, then that’s going to be the tell-tale sign. Meanwhile I think it’s still going to be another couple of years of figuring it out.

Do you think there’s a bit of reluctance or vagueness by some within the telecoms industry around Open RAN which might be confusing the issue?

It’s like anything when you moving into new technology. There are so many definitions of the of the word openness, that’s one concern. But the main concern is… OEMs are driving openness towards customers, and there are great examples of that – Rakuten obviously did that with Nokia, we’re seeing it with Dish, we’re seeing that with a number of different customers around the world – so there’s no question it’s happening. What will need to happen is that the CSPs can be satisfied on the performance and that there’ll be absolutely no changes that become service impacting, if they can be satisfied around that the logic for driving large scale openness is going to be absolutely apparent. And then the real issue isn’t in the performance of the network elements, It will be on the orchestration of those different services as you advance beyond 5G.

Another area that seems to struggle with its own definitions is edge computing. Sometimes there seems to be difficulty to agree on exactly what it is by some in the tech community. How would you define it, and why is it a big deal?

A simple definition for me is it’s anything outside of the traditional centralised data centres. And I think that could be a phone or an IoT device and anything in between. The disaggregation of the of the architecture means that storage and compute will be much more relative to the individual use cases, and then you will see – as the sophistication of the of the storage and compute happens at the edge –more and more and more advanced use cases happening.

But two elements need to happen. One is where do we see 5G? Where do we first see 5G chipsets? Not everything will have a 5G chipset of course so the edge discussion is really how does the user experience become transformed relative to different types of access technology, whether it be WiFi or whether it be unlicensed spectrum, or licenced spectrum? Again, there’ll be strategic considerations in there.

Then there’s the OEM consideration, and the application consideration which is of paramount concern. Where it does become synonymous with private 5G is when you have customers that will roam between private 5G networks and macro networks, where do they actually hold data? What are the security considerations for that edge location which could be included on site? Of course it could be an on premises consideration. And then what to what extent are they willing to have access to applications and data sets beyond edge activities as well?

So I think there are a lot of considerations around that. IBM’s role has become more and more essential in that especially around security. If you think about some of the access privileges, some of the types of data that are used, the fact that mobile networks and other types of access will be fused with operational technology, then you need to ensure that the edge play and the security play go much more hand in glove.

IBM is involved in lots of the technology areas that are kind of society and industry redefining.  With all the changes we’ve discussed coming to fruition, does the telecoms sector in 20 year’s time look very different?

I think it will. But there are universal truths that are out there that we can’t get around. Spectrum is a good example of that. It’s finite. So then the question is how much more can you squeeze out of it over time, and how much more efficient can you use it? And where the industry starts to become so much better in 20 years time is that it doesn’t fixate on one specific type of access technology or another. You’ll see marriage between satellite systems, licenced spectrum on the ground, unlicensed spectrum, and you’ll see that also with other micro layers of communication – device to device, you within the one metre range, not between the one KM plus range.

You’re going to see massive meshed interconnectivity, but again the industry needs to overcome some of the issues around standards, data security, those types of things. I think we’ve already seen the quantification of the industry. What we had is CSPs that were operating on a vertical strategy, I think you’ll see a lot more horizontal strategies as a result of that. And again, that started with firms like Vodafone breaking out and having Vantage Towers. And then you’ve seen the growth of firms like American Towers and Digital Bridge, they have a definite role to play.

The question is, will they have the same multi tenancy role or will they move into the active layer later on? And to what extent can regulation keep up with that? And then the platform players – actually the CSPs become platform players – the big question for them is what cloud strategies do they work on? Do they work on hybrid cloud strategies? I think everybody agrees that a hybrid cloud strategy is the right way forward, but how do they evolve their relationships so that the edge play becomes a real estate play?

Whoever’s got the service at the edge will be able to serve customers in unique and differentiated ways. So we are going to see a redistribution of the value. There are these notions of the death of the CSP and I think it’s silly. I don’t I just don’t see it like that. I think the regulatory environment would consider that, rightfully in my view, as anti competitive. And you will put too much resource in the hands of a of a few. We don’t want that as a sector. But we do want openness and I think when you see openness, you’ll see more players and when we see more players, you’ll see more innovation.

It’s a very, very interesting space. It’s complicated, which is why you need a lot of tech and technology and you also need a lot of consulting and systems integration expertise as well. The one thing that I think we also bring, and I think this is where we’re excited as well, is that there are not many firms in the world that really understand the telco sector and the way that the telco sector is evolving, and the industries which they are supporting as well.

If you look at IBM, we serve multiple industries. If you’re a CSP you’re serving things fairly universally, if you’re a web scale player you’ve also been serving things rather universally – we know intimately how the financing sector works, how the distribution sector works, so our ability to engage with our customers and stakeholders with insight and wisdom from those different markets and industry verticals is a pretty unique place to be.

Check out the other half of our interview with IBM, in which Rose claims AI is going to be central to telecoms going forward, but that ethically deploying it is key.

Get the latest news straight to your inbox. Register for the Telecoms.com newsletter here.