Andrew Hardy of NXP: IoT scale, Ultra-Wide-Band, and San Jose

Thanks, Pete. We look forward to a good, robust discussion today.

Yeah, no, I think it's a great question. It sort of spells out the relevance of the conversation we're about to have. I've been in the semiconductor industry for north of 20 years now. Spent the majority of my time in San Jose and had been involved in many different trends in the industry, starting off with the .com era and the networking spike that laid the infrastructure for a lot of the technologies that we leverage in our world today. And then moved into a lot of focus on the mobile side of our business and had been in that for a good four or five years. And then ultimately joined NXP about five and a half years ago, leading the Americas based in San Jose, where we've been for the last, well, north of, actually the last five years have been focused on developing the Internet of Things and edge computing. And the reason I brought up those prior trends was because in a lot of ways they took shape very much like the IoT market is right now, in that they started off with a lot of hype and a lot of buzz. And then ultimately, as companies got into the details behind the concepts and the visions of that particular technology, really a lot of the hard work started after that. And eventually the vision came to fruition, albeit at a slightly different timing than originally had anticipated. But I think to a large extent have had a very big impact on our worlds that we live in today. So looking forward to talking to you today about the IoT journey and where we see NXP today.

I was focused on the DSL market and central office DSLAM market.

Yeah, no, it's the million-dollar question, right? And if we look at, you know, going back to the sort of hype cycle, I think we have moved past the hype, so to speak. A lot of the companies that we talk to in the various different markets, and that includes retail, appliances for the home, home security, industrial, automotive, they all absolutely see the value proposition. And there are now proof points and examples where certain companies have come out and started to take advantage of the data that they're capturing on the edge and making the user experience significantly better than what it was before the technology was there. And even if you look at CES this year, a lot of the conversation moved from the vision into more of an implementation state. And so to me, it's now a testimony to the fact that people are deep into the details now. A lot of customers are going through all the various components of what are required to build a fleet of IoT devices. But what we've sort of uncovered in this is that the complexity that is there with building an IoT device in terms of the development of the family of products is substantially greater than what a lot of these customers had anticipated. Couple that with the fact now that you've got additional challenges of security that you have to manage to preserve your brand, especially as you enter into a home or an industrial manufacturing plant or in retail and you're dealing with people's data, that has become at the center of the discussion. And then in addition to that, people want to take advantage of machine learning and the vision and the potential that that technology has. And so you've added on two additional layers on top of what is the traditional, as you mentioned earlier, software, embedded software, power management, the price-to-performance ratios and so forth. You've got these additional layers of complexity that you have to manage. And so what we're finding is customers' development schedules are expanding because of this complexity. And so you start off with a business case that you're aiming for a market window, and as you get into that development, you find you can't either scale at the pace you need it or your execution isn't landing at the mark that you had. And so that's creating a problem. Now, as you and I have talked about for now a year or beyond, that problem, I think in a lot of ways, presents an opportunity for NXP and Microsoft to really go off and try to help solve this issue, to make that experience much more simple so that our mutual customers can invest the resources that they have on the things that will differentiate them versus having to build a new infrastructure.

Yeah, I know that's the ideal. I mean, I think, and also depends on the market, right? I think you guys sell into a lot of markets, the consumer market, the commercial market. I know for myself, like I get these AI cameras, these connected cameras for my house and stuff, and I had one of them that stopped working recently, and it's kind of almost disposable. It's like that's 30, 40 bucks. It just kind of died, so I'm like, huh, chuck in the garbage. But obviously, if you're like Chevron, you know, and you're outfitting all your oil rigs with all this tech, that stuff needs to be serviceable, you know, and it's got decades worth of stuff going on. You mentioned security, actually, and I know that NXP has this huge position in the market in security. A lot of people maybe don't know that. Can you elaborate a little bit on all the stuff you're doing around security and secure elements and like transit cards and stuff? I mean, the footprint of NXP is massive.

Right. So NXP, for over 15 years now, has been investing heavily in security. We've spent the majority of our time focused on the financial transactions business, for whether or not it's the chip on the credit card to authenticate. That eventually moved into the mobile phone, and then we took that business and developed the mobile transit business as well. So if you go through a metro across the world, more than likely you're going to be utilizing NXP's technology for getting access to the metro. And so after we established that market, and given our focus on IoT, we've taken this wealth of knowledge in security and started to apply it to the IoT market. Now, the challenge is you're dealing with a connected device on the edge. Obviously, that device is talking to the cloud. And so when we went to look at how we saw the security problem, it became very apparent to us that we had to have a very tight relationship with the cloud provider that put as much weight on security as we did. And so, you know, no pun intended, but this is exactly why we went to Microsoft, right? Because we felt that there was a clear heritage and importance in the design cycle that we saw that Microsoft put on their own products related to security. And I think that general philosophy of the importance of security married well with NXP. And so as we try to solve these security problems, which are different in IoT than what we've dealt with in the past, we feel like we have a good partner with Microsoft that has really put that at the core of the initiative that we're trying to solve with IoT.

Yeah, I mean, last point on that, you're absolutely right, but user experience has to be seamless as well. And I think that you bring up a really good point, in that if you put a lot of the responsibility on the user, they may not always actually take advantage of it. Sure. Building a solution and a service, frankly, and that's why we put a lot, we think Azure Sphere is a great bar to set for security, is there's a lot of seamlessness and service that goes behind that that customers no longer have to really worry about implementing a high level of security for their devices in the edge. And so we feel like that's a really good bar that's been set, and obviously intend on trying to raise it over time.

Yeah, I think you're going to start to see a lot of the machine learning capabilities that oftentimes we probably experience in the cloud today, you're going to see that starting to move down to the edge into smaller devices. And more and more these devices are going to be able to learn behaviors and start to become more intelligent with time. And so by moving that machine learning capability, which is trained in the cloud, down to the edge in a cost-effective way, is going to improve both the real-time nature of decision making, it's also going to reduce the cost of a lot of data going back and forth to the cloud, so it'll be done in a more economic fashion. And so ultimately, you know, the decision making of those devices is going to improve. You know, ultimately, also, I think from a connectivity perspective, more and more devices on the edge is creating bottlenecks on the edge. And so you're going to see new emerging wireless connectivity devices that will improve efficiency in the Wi-Fi networks. Obviously, 5G presents a significant opportunity for more intelligence in remote areas. And then new technologies like ultra-wideband that will ultimately provide more precision accuracy of actually identifying the location of devices and people, literally down to 10-centimeter spatial accuracy. So that's extremely exciting to think about, you know, identifying in a way that maintains privacy but also allows you to sort of walk seamlessly through your home or through your workplace and gain access to devices without having to connect to a certain network or log in and so forth. That can be done through a typical authentication of your device within a certain proximity and knowing it's you.

Yes. And the first area that this is going to start off is really in the automotive center for key access. And that particular technology is going to offer a more secure way of authenticating that you have the keys and where you are located, so over a pretty long range. So you can authenticate and pass keys along across mobile devices and so forth.

Well, actually, it'll just potentially allow you to eliminate your keys, so you can use your phone for that.

And then ultimately, though, then you can start to see that playing into accessing the home, or into an enterprise location, or detecting in an emergency where people are in a very close proximity. Right. So there's a lot of potential, even in retail as well, as people walk through the aisles of the supermarket and where they're spending time and not spending time, managing traffic.

You know, I think the next area that we have got a lot of investment going into right now is in the automotive sector. Basically, you can start to see that the automotive industry is starting to revolutionize with the electrification of automobiles, and there's the vision of autonomous vehicles as well. But the entire architecture of an automobile is really sort of being revolutionized right now. And it's almost, I mean, it really is frankly the ultimate edge computing platform, because it's on the edge and connected to the cloud, and there's a lot of latency components that are challenges that you have to deal with there. And so we see customers, and we are as well, really looking at that problem very differently than we had in the traditional automotive architectures. And it's almost a lot like a small data center on wheels, and the networking and the storage and the decision making through processing and how that's distributed throughout the vehicle is an area that's getting significant investment right now. And it's emerging new players as well, too, right? And I know Microsoft is in that conversation from the cloud side. We are absolutely in the discussion from secure edge gateways in the automobile as well as the infotainment side. So for us, it's I think just naturally over the next five, seven years, the work is being done now to come up with very new innovative vehicles that we will start to see driving around us in our world.

There's a lot of examples of that. I mean, ultimately, what's one of the biggest decision criteria for an automobile is a maintenance record, right? The ability to quickly diagnose and improve that experience, because ultimately there is going to be maintenance, but the ability to go in to get your car serviced, know precisely where the issue is when it occurred, and maybe even potentially have the components ordered for you prior to arrival, that just changes the complete user experience altogether, right?