ARM CEO Simon Segars talks strategy, smartphones and IoT

When you arrive at ARM’s Cambridge HQ it isn’t necessarily what you might expect of a global tech giant. To put it bluntly, it’s next to a field. One minute, cutting edge silicon design, the next minute, crops. But once you get to know the company you realise this is entirely appropriate. ARM has become a huge success precisely because it’s content to stay in the background, enabling other companies to hog the silicon limelight, so a healthy measure of British understatement is entirely appropriate.

ARM CEO Simon Segars is the personification of his company, not just because he’s the boss but because, like his predecessor Warren East, he is an entirely unpretentious, down-to-earth, straight-talking person. Even being based in the US west coast for many years doesn’t seem to have sent him down the hyperbolic path apparently considered obligatory by so many of his tech industry contemporaries.

It’s been around a year since Segars took over from East at the helm of ARM, so we commence by asking how that first year has gone. “It’s been great – I’ve really enjoyed it,” says Segars. “As you know I’ve been with the company for a very long time and helped define it for the last 20-odd years, but there is a difference between being part of the management team and actually being the CEO and it’s hard to really put your finger on what that’s going to be until you’re in the chair.

“I’ve spent my last 12 months talking to people in the business and outside it. You get to meet phenomenally interesting people; the business card gets you places you couldn’t ordinarily go. I was having lunch at Number 10 [Downing Street] the other week with a whole bunch of CEOs and I was really impressed. I reached the ‘shaking hands’ moment and David Cameron said: ‘Oh, this is Simon Segars and he runs this company called ARM, and their chips are in all these smartphones.’ He just knew that.”

But apart from meeting the great man himself, what have been the other highlights of the last 12 months? “The business has done great over the last year; we’ve had very strong licensing of the technology we’ve been working on for a long time, such as Cortex A,” says Segars. “We’ve been developing version 8 of the ARM architecture for a number of years now, where we introduced 64-bit processing. A lot of algorithms around networking, for example numerical algorithms that run in the cloud, all assume there’s a 64-bit number to play with, so it has enabled the ARM architecture to get used in many more places and it’s broadened the market for our customers.

“64-bit definitely has a role in mobile too. You look at high-end devices, they are approaching the limit of how much memory you can put in them, and the way we architected version 8 is very clean, very pure, so we’ve been able to produce a very efficient 64-bit architecture as well.”

The role, business model and products associated with ARM are typically contrasted to those of chip giant Intel. Where Intel makes money from manufacturing and selling its own hot, powerful processors, ARM sells its technology to enable others, such as Qualcomm and Mediatek, to make low-powered ships. As you would expect, Segars isn’t about to get drawn into directly addressing his competitors, but he’s happy enough to address the broader competitive environment.

“Our innovation is magnified by the innovations of our licensees so you get a very diverse range of products in the end-user space,” says Segars. “If you consider that against just one semiconductor company, regardless of the technology it’s based on, the more people you’ve got innovating around this common architecture the better.”

“A huge milestone for ARM and the ARM partnership was going through the 50 billion mark of ARM-based chips being sold by licensees, cumulatively. That is a phenomenal achievement and the rate has been increasing pretty comfortably. It was 10.4 billion ships last year and eight or so billion the year before that, and if you extrapolate the first half of this year we’re on for 12 or so billion this year, so around 20% growth year-on-year.”

Only half of the ARM-based chips shipped last year appeared in mobile phones, and this diversification is a cornerstone of ARM’s forward strategy. One trend ARM has been very keen to promote has been the Internet of Things (IoT), so we ask Segars to give us his view on how that’s progressing. “IoT is still in its very early days, but the combination of a chip, some wireless connectivity and a sensor, put together in a really small package, can be embedded in almost anything for almost no money,” says Segars.

“And then you create this enormous network and probably of interest to your readers is how the data that comes from that then gets into the network infrastructure. IoT is all about gathering that data and the money to be made from IoT is all about how you extract value from that data. But you’ve got to move it from where it’s sensed up into the cloud and obviously that’s where the network comes in.”

What next?

While any well-run company needs to plan for the future, that can’t be at the expense of paying the bills, so how does Segars achieve this balance? “When I took over from Warren, it’s not like ARM was broken, and having worked here for so long and been part of Warren’s team, what we’ve got is partly due to what I’ve done,” he says. “So it’s not like we’re suddenly going to move into making fizzy drinks and jeans just because I’m now in charge. I think the opportunity ahead of us is to grow large market shares in markets outside of mobile. Our 95% market share in mobile has been achieved over a long time, but all that success we’ve been having in other markets represents a big opportunity ahead of us.”

“If you look at mobile in terms of application processors, it’s a $13-14 billion silicon market today, growing to around $20 billion in the next five years. If you look at enterprise, which for us means network infrastructure and cloud servers, that again is worth about $13 billion of silicon today, growing to about $20 billion in five years’ time. Our targets for server share are quite modest, these things will take some time to ramp. I think we’ll see some first production systems this year, but probably grow to something like a 10-15% market share over the next five years.”

Looking to grab a 15% share of the massive server market from a standing start actually seems like a pretty aggressive target, so what makes Segars think this can be achieved? “Obviously the Internet already works fine, but it’s all based on a standard system architecture in the datacentre which, as workloads change in the cloud, doesn’t look optimal,” he says.

“If you take storage, for example, photos that get uploaded to social networking sites tend to get looked at for a relatively short period of time, but then stored forever. So if you’ve got some big server box sitting there managing a disk drive that’s not actually getting accessed very often, that’s very inefficient. There’s a much better solution to that, which is about integrating a processor that’s the right size for the task with the networking and the storage interface, and having something smaller, lighter and much lower powered.”

While grabbing some server share from Intel would obviously be satisfying, networking seems a more obvious market opportunity for ARM. “Networking is changing; with the growth of mobile devices you’re getting ever more data moved across the internet,” says Segars. “With the growth of IoT, again you get more data, and that data has very different characteristics. Systems are designed to support the volumes of data associated with smartphones, but with the Internet of Things there may be 100 times more devices that only have a very small amount of data very occasionally. That creates a very different load on the network, so the network is going to have to evolve. It requires customisation, scalability, it’s going to need to be quite flexible.

“And the third big growth opportunity for us is in embedded, where the really small, really low-power processors have been licensed really widely – we’ve got over 200 licensees of Cortex M now – to companies you’ve never heard of, who are off innovating in different ways looking at this IoT opportunity, and that’s why the volumes are growing to 900 million in the last quarter. It’s going to be really interesting to watch that innovation play out, and again from a silicon perspective each chip might not cost very much, but it’s around a 13 billion dollar silicon market today, moving to around 20 billion in the next five years.

An easier way to keep the share price moving in the right direction might be to just charge more for licences and royalties. After all, what’s the point in dominating a market if you can’t exploit it, right? “A very common question I’ve had in the last 12 months is: you’ve got such a high market share in mobile, why don’t you just quadruple your royalties?” says Segars. “We could, and it would generate more money in the short term, but that’s not the way to build a really long-term business.”

“We’ve got 95% market share in mobile and there’s no reason why we can’t have similarly large market shares in these other markets, but if our customers think ARM’s just fleecing everyone it’s never going to happen. So, we’re about volume, we’re about growth, we’re about total size of the market and evolving into opportunities in the future that nobody can think of today. It seemed ridiculous that ARM processors would be in servers some time ago, but our licensees are now looking at that as an opportunity.”

The state of the industry

As the CEO of a company that partners with so many other tech players, Segars has a privileged perspective. So how does he view the current state of play in the mobile industry? “I think we’re at quite an early stage of LTE, globally,” says Segars. “Things do go in cycles, so when 3G rolled out you had handsets that could make phone calls and send SMSs really quickly, and it took a while for applications to develop that really utilised the relatively high performance of the network.”

“And I think we’re also in a period where there is experimentation going on in terms of what other sensors you put in a phone and what they talk to. At the moment it’s relatively easy to put a bunch of sensors in a phone, but there’s not much for them to talk to. I think you’re going to see an evolution of the things they’re going to talk to, such that ultimately this device that you carry around is the controller for everything else you’re interacting with.”

“There might be location-based services that spring up because you walked into your local supermarket that you go to every week and suddenly the phone realises that and then the supermarket will present offers based on your personal shopping history. The handset is going to evolve such that it’s very personal to you.”

Talking of personal, the hot new mobile category right now is wearables, especially since Apple decided to get involved. And, as you might expect, Segars is a fan. “Wearables, again, are at a very early stage,” he says. “To me this [gestures to the Pebble smartwatch he’s wearing] is really useful. For this to beep when my next meeting is about to happen, or buzz on my wrist when there’s a phone call coming in because by phone’s in the bag on silent, I find incredibly useful, and I do genuinely wear it every day.”

“I appreciate I don’t represent a typical consumer when it comes to adoption of new technology, but these things are going to find a use that does resonate with a lot of people, and I think health is going to be a huge area for wearables. The medical systems of the world are geared up to deal with acute illness – you go in, your leg’s hanging off, please fix it – but as the population gets older and less healthy chronic illness is the problem, and managing that is a huge issue. So to track your wellbeing and to have communication going on with your healthcare provider, such that if you need to come in they’re going to tell you, and there’s a bunch of data that’s really relevant about your state of health, all represents a real opportunity.”

“On the networking side, with the evolution of the network topology, you’re going to get small cells, wifi offload, and that drives more embedded processing, so there’s a big silicon growth opportunity there for us as well.”

“People might not realise there is some growing momentum behind the deployment of ARM-based silicon into networking equipment. As people look at NFV and SDN and the evolution of the network and network topology I think there’s a huge place for ARM in that. We’ve been getting a lot of interest from networking vendors and it’s because they look at ARM’s business model and think they’d like to do software development, but have choice in supplier.”

“If they focus all their software expertise around one architecture, then they’re making a major long-term commitment to it. If you do that with ARM you can get chips from a number of suppliers in this space, which has big benefits. Everything that made us successful in mobile turned out to be useful in lots of other places.”

There’s that ARM business model again – doing the hard work on the silicon side so other tech players can focus on what they do best, whatever that may be. To conclude we ask Segars to give us his parting thoughts on the industry his company plays such a prominent, if discreet, role in.

“It’s easy to get bogged down by the difficulty of transistor manufacture and so on, but when I look at the opportunities for deployment of chips with processors in them I can’t see any slowing down of the demand for silicon in my career,” says Segars.

“Also, because of the innovation drive coming from things like kickstarter and the availability of all these chips, we’re going to see a phenomenal amount of innovation, and it is just down to what ideas people can come up with and how quickly they can deploy them. So I’m very optimistic about the next few years.”