We spoke to Adam MacHale, Cisco’s Vice President – Service Provider EMEA, to get his thoughts on 5G, edge computing, VR and AR, and who is going to pay for it all.
MWC is just around the corner, and the current state of all the big telecom talking points – Open RAN 5G, Edge, the metaverse, IoT and more – will be chewed over in the show halls and on the stages. Ahead of the show, we spoke to Adam MacHale, Vice President – Service Provider EMEA at Cisco, to pick his brains on some of these themes, and crucially how the flashy trials that are used to demonstrate the technology can be converted into functional and profitable businesses.
What are going to be you showing off or talking about at MWC this year?
Hybrid work is a huge area of focus for us. It’s going in different directions – as we came out the pandemic lot of companies were very distributed, to [then] come back to the centre, and obviously the solutions around that. Sustainability is a link I think you’ll hear across everything… that links nicely in with hybrid work and how people are approaching that and what they can do.
It’s interesting what we can do around building technology, and what we can do around the distributed workforce and getting that right. And I think that’s still something that’s really developing. Security across our whole portfolio and how we can integrate, just the size of the attack surface, and the need to react extremely quickly, rather than having security around the edge.
In the telco space, my space in particular, a lot about what we’re doing with Silicon One which is the development in our own chipset and some of the optimization and performance that we can do around that. We’re doing a lot around optical and bringing the layers together. So simplifying down the layers, which again links in nicely with things like sustainability – a lot less boxes – but also using less power and it’s less of a complex infrastructure.
There’s a big conversation going on at the minute – everybody recognises that the public cloud is really important and everybody’s workloads [are] there. But the sovereignty question has really been called back again.
That’s where perhaps you traditionally see Cisco playing. I think there’s a lot around what we can do within the data centre. There’s a big conversation going on at the minute – everybody recognises that the public cloud is really important and everybody’s workloads [are] there. But the sovereignty question has really been called back again, I think everybody’s now reviewing that you need a balanced approach. Some workloads may be local, some may be remote. So that’s another sort of big area of work for us going forward.
And then if we look into mobility we did a press announcement with T Mobile in the US around 5G SA, where they’ve got an excess of 200 million subscribers on it. I’ve seen some of the commentary and I’ve got views on how well 5G is going and perhaps what we need to do within that space.
From Cisco IoT, years ago we made the Jasper acquisition that’s now called Control Centre. We’ve got over 220 million Sims on that platform. And we see that as an important element of profitability for the telcos going into the future. Not easy, an extremely overhyped area in the market, but we are having successes in that space and that’s something that we are building on.
You mentioned that you have some views on how well 5G is going. So, how well is 5G going?
I think there’s a general feeling in the industry that it hasn’t delivered on the business case and perhaps and some of the hype that was laid out. There have been huge investments, I think it’s going different in different regions around the world and how that’s progressing. But the magic use case that everybody’s been looking for has been a struggle.
I think in a consumer space, yes, it definitely has enhanced the throughput when you can get good 5G coverage. It definitely enhances that but I don’t think it’s increased the average revenue per user. In some of the markets in the Americas and APJ, they’re starting to see some areas… there’s been some anecdotal conversations around ARPU increases recently, but for me it was a big technology play out there in the market, and we need to really focus on the value that the telcos can provide and how can they use that to go forward.
I think there’s a general feeling in the industry that it hasn’t delivered on the business case and perhaps and some of the hype that was laid out.
And that links back into IoT and some of the areas that we need to move into, it’s probably not more consumer based connectivity. That’s the baseline you have to have. And I think we need to probably as an industry do a better job in things like rural locations still, and make sure that everybody has access. We know the positive impact that that can have on rural communities for example. So we need to keep focusing on making sure that there’s connectivity there, but I think IoT and the supplement of the b2b connectivity – fixed wireless access – is a big push where the spectrum availability for example, in the Americas, and I think there’s things that we could be doing better within that space.
But the telcos – whatever term we want to use – they need to continue to move up that value stack. How can they get more engaged, open up their platforms? I think a big part of that is opening up API’s and allowing third party developers to develop on their platforms to be successful. It’s an easy thing to say, it’s not an easy area to get into. But I think that’s going to be critical as we go forward.
There doesn’t seem to be an easy answer to how telcos can make a better return on it, which would seem to be acknowledged by their claims that big tech should start helping out. Do you think that’s sustainable in and is there a solution to it?
It’s a really complicated area of regulation. And it’s something where governments need to work with telcos and with vendors and the end customers to get that right. There’s many different aspects that – if we over regulate, we stifle innovation. That’s often one of the conversation points, but we do need to make sure that the business models are there to be successful for the telcos to go out and do these.
I think it’s an interesting case study around what happened in copper unbundling. I know that’s looking way, way back – but the copper unbundling happened and then fibre take up didn’t really happen as quickly as anybody would have liked. And it’s only during and after the pandemic that there was a huge investment in fibre, which is really important. But some of the regulation around that and a protection for who makes the investment for the fibre in the ground, actually realises the benefit before that’s opened up.
I know that the telcos are lobbying with the hyper scalars around who uses the most bandwidth… that whole area I think has been stable for a long while, but is it something that should be looked at? Absolutely.
So there is a case for it. It’s going to be a complex discussion – I know that the telcos are lobbying with the hyper scalars around who uses the most bandwidth. I think there’s an interesting sustainability piece to building out this infrastructure and the value of the traffic – and that links in with net neutrality. That whole area I think has been stable for a long while, but is it something that should be looked at? Absolutely.
With the recent emphasis on Open RAN, how much more involved are you in in setting up mobile infrastructure up than you used to be?
We were probably more in radio a few years ago, we’ve made acquisitions in that space. We did a lot around small cell and what can be done in building. It’s a it’s a difficult space in the radio space, it’s a complex area and I think that’s why there are some very strong players within that market. Cisco is extremely strong in Wi Fi and in the enterprise space, which has been extremely successful for us. But to your question, around Open RAN, I can see the reasons why it’s really important to look at that space and to open up the ecosystem to get more players within that space, which will help drive innovation.
From a Cisco perspective, we always try to be as open as you can. If you think about the success of the internet and everybody that’s played within that space, it’s a very open environment. And perhaps the radio ecosystem has been pretty closed. It’s slowly opened up as we’ve gone through the generations, but it’s still fairly closed. I think what we saw in the router space with disaggregation – where people wanted to buy the hardware from one place and the software from somewhere else, and maybe even get a different chipset – If you had asked us about that five years ago, we probably would have had a fairly closed architecture.
It might lower the cost of the hardware, but the integration costs – because you’ve now exploded that end device – may be significantly more.
Whereas now with Silicon One, you can buy our silicon, you can buy our software and run it on a third party or you can go with the whole stack. I think that needs to come to the radio space. The challenge is that’s going to take time and it probably isn’t moving quickly enough within that space. One example is if we if we do that in a very open environment running on x86 It won’t be more power efficient. And with the whole sustainability conversation and what we need to do to make sure that that’s been done in the right space, we might drive things in the wrong direction.
So although we’ll open up the architecture and allow for a lot of innovation, it may not meet the aims of what’s trying to be done. I think cost is another thing that’s been brought up, that this will lower, lower the cost of the hardware. It might lower the cost of the hardware, but the integration costs – because you’ve now exploded that end device – may be significantly more so you may be saving on the hardware but then you need to be able to operate.
We’ve done successful deployments in this space working with partners like Rakutan, we have a good partnership with NEC. So we have good visibility, where we’re more involved is on the infrastructure piece because that brings in some interesting latency and throughput requirements out of the edge. The distribution of the data centre, because now we need to run workloads much closer out towards the edge and that changing the architecture.
But are we in Open RAN from the Cisco perspective in the radio? No, that’s not the space that we’re currently playing. Okay. Where we are playing is in private 5G, and we have partners who are using some of the Open RAN technologies for radio so we’re partnering with perhaps some of the smaller radio vendors within the private 5G space.
Moving on to another telecoms buzzword – virtualization. Could you give us a real high level view of how it’s being applied to the telecoms industry, and what the what the benefits are?
I’ve been on this journey for many years now. A simplistic answer rather than me just jumping into the technology, the one I’ve seen used many times before is like a smartphone – and the ability for you to be able to run multiple applications rather than buying your GPS system in your car that you stick on the windscreen versus a compass – so combining all those things was enabled by software.
And that’s what we wanted to be able to do with the telco infrastructure. Because typically, you would have to buy many different boxes, all bespoke, all running a different operating system and software on top. So what we’ve done is taken the software components of those boxes, and now we can run them a bit like applications in a cloud and that gives us the flexibility to switch them on, to expand where we need, to contract where we need, to do software based upgrades, rather than it being intricately linked with the hardware.
What we’ve done is taken the software components of those boxes, and now we can run them a bit like applications in a cloud and that gives us the flexibility to switch them on, to expand where we need, to contract where we need.
That level of flexibility will allow the telcos to be much more agile. And initially we used to take, if you like, the computer operating system and run it inside a virtual computer. But now we’re writing the applications to be containerized and highly efficient, which allows us to get much better performance. And I think part of that conversation is it’s relatively easy to take the software from a box and run it virtual, but what you tend to find is it runs very inefficiently, which isn’t a good sustainability message.
What we need to be able to do is make sure we can run it highly efficiently and still have the advantages of software. And so that’s the journey we’ve been on. Now most of the workloads within the infrastructure with the exception of the radio are virtualized, but we’re starting to move towards things like bare metal, where you can run it directly on the hardware. The T-Mobile example that I mentioned, that’s all running now on bare metal and they estimate they’re getting a 20% efficiency over running a virtualized.
So a lot of this is about efficiencies, bespoke solutions…
It’s the agility. Whereas before you’d plan a deployment of 3G, and you’d spend two or three years saying, we’re going to put the boxes here, we’ve got this capacity, you do all your capacity planning, and then you’d go for it. Now you need to launch a service, try it for a month, and if it’s been successful, give it more resources. And if it’s not, you want to switch it off. So that agility, bringing that to the telco space is going to be important.
Cisco is also big into edge computing, could you give us an explanation of edge computing, and what are some actual examples of it being used?
Today when you’re consuming a service on the internet, and if you think of the concept of a cloud, often those services are extremely centralised. A few years ago, but still in some cases, a whole country’s mobile network might peer in two or four locations. So as soon as you got on the internet, you’re always going back to these hub locations in the country. Now the good thing is you get scale and efficiency. The poor thing is the latency, the round trip time – even just the speed of light through the fibres, is significant.
And if you want locally based services, it’s much more efficient for them to be nearer to you. So you want to consume it near the edge. So when we talk about the edge, we’re actually talking about getting it nearer to the end user. If we take the example of augmented reality, where you’re overlaying things on top of what you’re seeing, you need low latency to effectively be able to do that – your body your eyes are very sensitive to that correlation. There’s many, many different use cases that you can think of to be able to do that, pushing that closer out to the edge.
The advantage for the telcos is they’re the only ones that can do this because they own the infrastructure.
The disadvantage is it becomes more complex because instead of doing four sites now we’ve got 1000 sites or 10,000 sites, so you need to be able to manage that. And because you don’t have scale, those services are probably going to be more expensive to implement. So you don’t want to do it for everything. One of the cases that’s been around for many years, but I think he’s seeing very rapid adoption is smart caching. If you think about all the streaming services, with Apple TV, Netflix, Disney plus, we have a solution with Qwilt that’s being pushed right out to the edge with many European operators, that when you stream, instead of us streaming over the whole infrastructure, you’re only streaming over the last piece, which reduces significantly the amount of traffic that needs to be carried. So it makes those networks much more efficient. But that’s the site where we can host security services, or collaboration services or other things to consume where there’s low latency.
The advantage for the telcos is they’re the only ones that can do this because they own the infrastructure. There’s a lot of interest in how can ‘over the tops’ host within those environments, and everybody else wants to host in those environments. it’s going to be a real important initiative going forward. The latency question is sometimes overhyped. WiFi is very low latency and you can put a service very close to it for doing something within a factory floor. But When you’re looking at 5G services in the wide area, those low latency services with edge compute could be could be important.
You mentioned WiFi there, which would seem to have much more to do with virtual reality, because edge computing combined with 5G is about being outside, not being in your home. I just I don’t see that people would be wearing virtual reality headsets anywhere other than their front rooms. So I wonder how that plays into the picture, and if there are any other applications for it?
And this is where we just to your point about the user experience, and Meta… people say ‘we need 5G for meta’…well you’re connected by WiFi and you’re indoors and, so you don’t need 5G necessarily to run Meta or any of these other pieces. I think augmented reality, where you’re tagging things outside over what you’re doing… think about sports venues. You can hold up the phone it will tag the players and tell you who they are, so you get an augmented reality experience straight off your phone, but you’re outdoors, your live, you might be using WiFI but you might be using 5G.
But the latency of that is really important because there’s a lot going on in front of you. There’s a lot of data processing going on. If you click on a player and say I want to watch this player they can track that for you throughout the whole event. And that’s something again, that where we’re working on with Qwilt because that means you may have 20 camera feed, and you can choose which view that you want. And you want to do that much more locally to that event rather than having to go much further away because the amount of processing.
People say ‘we need 5G for meta’…well you’re connected by WiFi and you’re indoors and, so you don’t need 5G necessarily to run Meta or any of these other pieces.
So I think we need to be really specific about the use cases and careful with it. There’s not one that’s going to going to hit everything. But I think it’ll be some of those high value use cases in industry, and there are a lot of trials that have gone on. WiFi inside a building is a super effective technology because you’re not trying to provide the coverage from outside. But as soon as you start talking about airports, where there’s you’re going inside or outside, where you want to track containers in a boat all of these sorts of things, 5G and IoT – and we say 5G But it’s 4G as well and all the different Gs that will contribute to that – there will be some very high value use cases. Asset tracking – when you’ve got extremely expensive assets, the business cases is an extremely good one if you can do it effectively.
It’s technically possible to do the things you’re describing in a stadium of course, but do you think the general public has shown any signs that they’re willing to pay more for things like that? And if not, who does pay for it? When does it become a business as opposed to a trial?
I think we’ve got good experience in the connected stadium. I mean, people pay a lot, it can be thousands of euros for a ticket to go to a stadium for an event. How do you pick up that customer as they come into the carpark… how do they get that experience? And if they’re spending that much for a ticket, would they be prepared to spend another €20 or €50 to have the right sort of experience? They want to live stream a moment and we say look, we’ll make sure that you’ve got the right level of performance, but you might need to pay additional for that. In the right way, because we come back to net neutrality and doing that the right way within the stadium.
But I think the investment that has gone into stadiums, and we’ve seen it on Wi Fi in the coverage models… in the World Cup they had sensors in the football and they were streaming telemetry from the football… you need really good coverage within that.
Your question was around business models. I don’t I don’t know if consumers will pay more. I think that’s the challenge.
But I think your question was around business models. I don’t I don’t know if consumers will pay more. I think that’s the challenge. Maybe they’ll be some new gaming… we know that gamers will pay more for a low latency service that might give them an advantage in the home, but that’s not 5G. That’s going to be over fibre and where that’s hosted. But it’s the vertical markets… connected farming – when you look at the cost of fertilisers and you know the cost of seeds and everything, some of the clever things that can be done with connectivity just to make sure that they are doing that in the right way, the technology now has become so smart within that space.
But it only works if they have effective coverage on the farms and that tends to be in rural locations. So how can we provide that? Because the business case for that is extremely strong. Industry; we saw it in the pandemic when people couldn’t go into the factories… increasing automation within those environments, and those types of connected machinery which cost millions and millions of dollars, and downtime of 30 minutes can be hugely impactful.
So getting the connectivity right so you can run real time to some of those devices… that’s why I use perhaps the umbrella term of IoT, but not a tag on your bag so you don’t lose it – which is interesting – but it’s really these really high end business impacts. What we need, and this is one of the things that Cisco brings – we play a lot within those vertical markets. And the telcos don’t often understand those markets. So if we can bring them together in the right way, the business case can be extremely powerful. And for that you need a different, agile way of engaging.
I think I saw you’d written about ‘telco to tech-co’… and I think that is a really important shift. And I’m going to completely off on a different one here; skillset is holding us up within that environment. In many of the projects, the technology looks great, but there simply aren’t the skill sets both to deploy it, or the end users to actually be able to consume it and do something with it. So we’re doing Network Academy within that space, and we’ve got millions of students who’ve gone through that, but we did an announcement with ITU a couple of years ago on ‘how can we drive this skill set uptake in both developed and developing markets?’