Simon says: ARM is well positioned for the internet of things

[protected-iframe id=”273e8528e4aa1b7d54fb0173fdf33ce4-14960843-33105277″ info=”″ width=”640″ height=”360″ frameborder=”0″ scrolling=”no”]
Transcription details:
Input sound file:
10-16 am Session 2_1004.MP3

Transcription results:
Session name: Designing the Foundation of the Future of Mobile
Chris Albrecht
Stacey Higginbotham
Simon Segars

Chris Albrecht 00:02
We’ve got a pretty fun morning so far, right? But now stuff gets real. We’re going to bring out our next guest. We’re going to talk about “Designing the Foundation of the Future of Mobile.” We’re going to bring Stacey Hagginbotham back to the stage and she’s going to be talking with Simon Segars the CEO of ARM. Please welcome our nest panel to the stage.
Stacey Higginbotham 00:28
I think that was Chris’ subtle way of saying that I’ve been on stage a little too much today. I’m back because I’m so excited to be here. Right now I have Simone Segars, the CEO of ARM. For those of you guys who know me, you know I’m a huge chip-head. I fundamentally believe that everything we’re doing can be traced back to innovations of silicon because it can. I was thrilled to have you here and I thought we would start by piggy-backing on the previous panel that I had done. Talk about how IOT a reality from perspective because ARM has gone from smartphones, embedded inside many game consoles and set top boxes. Now you’re moving all the way in to servers. You guys did a networking announcement today with Freescale. You’re everywhere – the internet of things and you’re going to be all over it. So how are you guys working to make this a reality, other than providing the intellectual property that’s powering these chips.
Simon Segars 01:39
Thanks Stacey and thanks for inviting me to be here today. Good morning everyone. So as you say, we’re in a lot of places. The part of the beauty of all business model is we get an amplification of what we do through this big partnership that we’ve developed over the years, who have been looking and taking this based technology that we make and putting all of their cool stuff around it. And working with the ecosystem that they developed, that we developed and we, together developed, to actually make some of these products a reality. The panel that you moderated earlier was really interesting because some of those areas about privacy, security, standards – we’re spending a lot of time scratching our heads about that as well.
Simon Segars 02:20
There’s many thorny issues, which is, if it can be solved – hopefully it’s a when, but I think there’s a little bit of “if” in there. I think the internet of things could turn into something really useful and really interesting, but there’s some big barriers in that. I mean, you still touch on security. There’s data around us that you really have to worry about – who gets access to it, who owns it. If you want to engage with your insurance company, it’s a slightly different way. Maybe they’ll give you a set of bathroom scales and have you weigh yourself every morning and keeping your weight in the band, they’ll give you a better deal in your insurance. But how will they know that it’s you who’s using it?
Stacey Higginbotham 03:09
I would totally plan someone skinnier – healthier.
Simon Segars 03:13
I have a Fitbit, I stick it on the dog and kick him out of the backdoor at night and up the data, how do they know that it’s me using it? There’s a whole lot of authentication issues. Is this the device that they want you to use? Are you the person using it? Can you rely on the data? And what are they going to do with the data? What if I sell it to somebody else, then, who owns it?
Stacey Higginbotham 03:37
Now, that you’ve scared me utterly, we’re never going to get there. How are you guys doing this? I know that you’ve got partnerships with a lot of companies. You’ve got Xively. You’ve got security partnership with Gemalto and G & D.
Simon Segars 03:55
Giesecke & Devrient.
Stacey Higginbotham 03:57
I’m so glad you’re here. So you’ve got these and I’m wondering about the ARM licensing model, does that play into your willingness or your ability to build these partnerships as opposed to selling a chip? Does that make you think bigger and think more ecosystem-like?
Simon Segars 04:23
Yeah. I think, through our business model, we do have the potential of the technology to go in lots of places. There’s nothing wrong with building chips and I’m thankful that a lot of people to it. But if you’re going to do that, you have to focus on a relatively narrow set of application areas. I think, what we’ve done well over the years is look at how different application are growing and boil the requirements of those, down to a relatively small number of things that we can go and focus on. And then spend effort on developing the ecosystems to make them successful in a particular market, whether that’s joining a standard body that’s all around automotive or whether it’s developing a security solution. These things help the deployment of ARM technology and ultimately, help volumes for us, which is what we like. Security, is a great example.
Simon Segars 05:13
We started looking at security in the late 90s, early 2000s, thinking about mp3 and DRM because clearly, everyone is going to want that. It turned out that time, that they wanted DRM and it got in the way. We developed some security hooks into our processor architecture, to allow the processor to go into much more secure mode. We we’re thinking that it’s great that we’ve done the hardware and joked about somebody’s else will worry about the software. That was only true to a limited degree, so we started thinking about software and it became very fragmented. The joint venture that we put together about a year ago with Gemalto and G &D, is an attempt to remove some of that fragmentation. It allows people to focus on IPs and into more sophisticated software platforms to allow services to run on top, not on top of this very secure piece of hardware underneath.
Stacey Higginbotham 06:15
Is it important than to get the ARM architecture in a lot of things, but there’s also MIPS Space architectures and there’s XAD-6 architecture, is it important to have all of those things be the same? Like have all those people be a player in this security effort or can you do the developers and application people to build three sets of hooks or is that a problem?
Simon Segars 06:42
There’s a standard open platform, which sole defines an API into a high-level software, so then, there’s the implementation of that underneath. There are different ways to skin that cay. It doesn’t have to just be on-based.
Stacey Higginbotham 06:57
Kicking a dog and skinning a cat, I’m a little worried about your pets here. Let’s take it beyond the internet of things. Let’s talk about smartphones because this is Mobilize. I don’t just care about the internet of things and you might not realize this. You were in smartphones from the beginning and everyone has the ARM architecture or a license to mess with it. So what are you seeing coming down the road on what is our smartphone’s going to be able to do in five years? Or maybe it’s a watch, I don’t know, what is that going to look like or everything?
Simon Segars 07:38
My crystal ball is not perfect, but when I think about what the technology can do, I think the smartphone are becoming peoples most preferred way of accessing the internet and all its glory. When the internet of things interacts with people, the world becomes the interface with which you interact with the internet of things. The touchscreen is the primary user interface, I think that user interface will move. I think there’s a lot of sensors in smartphones and tablets, which allows you to interact with it in different ways – through voice, through gesture, whether you look like you’re having a bad day or not. Actually somebody mentioned Fjord in the panel the other day. These things will go from Fjord, talking about how they can build sensors in the car that could tell if you’re a bit stressed and then maybe turn the music down or not let the phone ring so you can concentrate on driving. These are examples on how technology can infer a lot more about what’s going on and I think you’re going to see that in phones. You can see a lot more sensors interacting with things, devices and interacting with you in different ways.
Stacey Higginbotham 08:53
If I have sensors everywhere and maybe my car tells me, like Audi, it talks to me and it gives me emails and that sort of thing, what will my phone eventually do for me? Does it just sit in my pocket and every now and then, I pull it out because I need the actual internet? Or am I going to not really need a phone because my information’s coming through Google Glass or some sort of an interface like that? Does it become a weird computer that powers all or ties in to all of these other things in my pocket?
Simon Segars 09:28
It depends on how you take it. I think it also depends on the performance of their networks and particularly the latency. In a world where you have zero latency and a ton of bandwidth and infinite cloud computing, then all your data can live somewhere else. Your whole experience can be in the cloud. The thing that you carry around where it just happens to be something that lets you see into it and if you lose it, then you just get another one. The reality though is – for a very, very long time to come, there’s going to be some tradeoff between what you compute on the device you carry and what you compute in the cloud. And that’s one of the tradeoffs that may change, but at the moment – for the foreseeable future, there doesn’t seemed to be a shortage of demand for compute performance in the phone. Be it raw CPU performance. Be it GPU performance or video streaming, all these kinds of things are going to rely on a lot of technology in the phone.
Stacey Higginbotham 10:29
What about battery life? ARM is famous for low power processors, but still, no one is happy with battery life. As I start putting more sensors out there and more computers in weird places, where are our advances and power coming from? When are we going to solve that or go there? Or is that essentially solved?
Chris Albrecht 10:59
I don’t think it’s solved. It’ll be great if one day batteries suddenly got a lot better, but we’re not banking on that. If you look at the trends over time, compute performance gets exponential and actually, it gets better in about 10% a year – it’s a growing gap. We are not relying on any fundamental shift in battery technology. It’s got to be about how you design the chips or how you put the silicons together and how you get really smart about activating the part of the chip that’s going to do job right now. And how you evolve computing, so you’ve got the right compute engine doing the right kind of task to do it the most appropriate way. That’s an intellectually challenging problem for us – solving the material side and battery–
Stacey Higginbotham 11:46
It’s someone else’s job.
Simon Segars 11:48
Stacey Higginbotham 11:48
What about things like energy harvesting chips, not necessarily for you phone, but for your sensors? That’s kind of a hardware thing that you might have to deal with.
Simon Segars 11:58
Yes, that’s something that we’re doing necessarily. Some of our guys talk to some of the researchers who are working on that because in the internet of things, it could be lots of things for lots of different people. One category is definitely very small processors with sensors and with wireless connectivity. It is placed somewhere that you never want to have to go and touch again. You don’t want to have to go and change your battery. You might not easily be able to get electricity to it. So energy harvesting is going to help deliver that. Again, it’s one of those things I’m hoping, that somebody else solves.
Stacey Higginbotham 12:31
One of the cool things that I’m really excited about, vigorous computing has made hardware cool again. We have all these companies here and they’re all making physical products. Obviously this is good for you because you’ve got more physical products and more chips sold. Can you talk about your thoughts on this idea that hardware is cool again? And how long will that last? How does that change business models? Or what you might advice someone to go do?
Simon Segars 13:08
One of the things that’s enabling hardware to be cool is some of the fundamental building blocks have already been put in place by somebody else and they really are inexpensive. You can buy micro-controller with a ton of functionality built into it for 50 cents. Then spend your time building a hardware product around it. You can program it easily. You can go online and you can find a ton of coding examples to have access or DIO, that’s in this micro-controller. The wheel has been invented, there’s no point in having another one. Unfortunately, people are putting a lot of the building blocks in place, which just saves people a lot of time. They don’t have to worry about owning a Fab, design the silicon, build the chip and now I can start working on my product.
Simon Segars 13:56
It’s all there, which is really cool. We have a program in ARM called “Embed,” which is about taking somebody’s on-base micros and putting them on a really inexpensive boards and providing a cloud-based software development tools. So for people who almost have no money can start working on a product. I think that helps unleash the creativity of lots and lots of people in the same way like the iTune Store, enabled software developers monetize their work really easily. I think, Embed are doing those things like that and Raspberry Pi, are enabling people to go, “Jesus, it’s a very useful platform. That’s really cheap. What can I do with it?” So I think it’s just accelerating creativity, which is great.
Stacey Higginbotham 14:44
What are the implications of that for start up or business? I look at Kickstarter, for example and I see this things and I’m like, “That’s such a neat product.” Then I think, is that a business? It matters to a certain extent, maybe it’s a lifestyle business as opposed to like you’re going to make the next iPod. How does this cheap availability of hardware and the ability to make something cool that you think about that you want to see in the world? What does that mean for an entrepreneurial?
Simon Segars 15:18
Hopefully, it helps fuel it. If the cost barrier to starting your project is lower, more people will do it and they’re still going bankrupt if it goes wrong. So I think, it’s going to drive creativity. It’s going to drive people to go and experiment. Some of the stuff in Kickstarter is really cool. You can see that people are buying Embed boards and the old boards and just putting things together in their bedrooms and if it works and it’s interesting, they’re going to raise some money – and Kickstarter is a great way to enable it somehow as well. So I think it just enables more people to participate and you don’t need masses of scale, you don’t need masses of funding to go and build something out of this pebble watch, which is fantastic. When that happened without Kickstarter – when that happens, there’s a lot of low cost electronics out there and if more people build products like that, all the better.
Stacey Higginbotham 16:10
More people are definitely building watches.
Simon Segars 16:13
I’ve noticed.
Stacey Higginbotham 16:16
I think I can have a new one every day of the week. I can’t let you leave without referencing the facts that it may not your biggest rival because it’s a very different business model. You guys and other architectures are in a lot, like Raspberry Pi are doing boards. Those are out there and here comes Intel, just a couple of weeks ago, releasing the Mini-Boards and the Galileo Boards. And now, they got a low power cork architecture for the internet of things. In so many ways, it must be very flattering. How do you view that Intel is basically coming in to your market?
Simon Segars 17:00
I think the market is huge. It almost goes to over 300 licensees about technology. People are building a whole wide-range of chips. The demand for silicon is only going to go up. So the market is big, it will support lots of people being in it. We think it’s a big benefit in a lot of the software we use that comes out of the ARM ecosystem. The market is big and it’s also very fragmented. IOT can be watches, it can be industrial controllers, talking to GEs back-end and it can be sensors and light bulbs. A vast of advanced and broad array of solutions that are going to be required for all of that. There’s no “one-size-fits-all” solution to it. Competition will ultimately drive the best solutions for it or another semi-conductor company is working on that, rather they worked on ARM, but I can understand that at the same time.
Stacey Higginbotham 17:53
When will Intel get the ARM license? Probably, no. They sealed that off.
Simon Segars 17:58
They still made some space on ARM over license.
Stacey Higginbotham 18:00
Very good. Simon, thank you so much for coming today. Every one, Simon.
Simon Segars 18:07
Thank you very much.
Chris Albrecht 18:14
Thank you Stacey.