Now, Lars, welcome. Welcome to the Aurora Exchange. Thanks so much for taking the time. So, can you do like a very quick introduction and talk about yourself, your background, and NXP, and what it does?
Yeah, so I'm a 55-year-old physicist, German father of three teenagers. Next to my physics, I did an MBA at London Business School. With that combination, I worked myself into the CTO role of one of the biggest European semiconductor players: NXP, the former Philips Semiconductors, that acquired 10 years ago the former Motorola Semiconductors. With that combination, we are one of the electronics players for all sorts of intelligent systems, robotics, big chunk in automotive, as well.
Does that happen like after also all these players first spun off the semiconductor parts, the semiconductor departments, and then there was like you who did the partial consolidation? Is that—
So, around 2000-2005, there was the understanding that the semiconductor branches of the big corporations are not earning enough money, and they are too volatile. And therefore, everyone, Siemens kicked out its branch, called it Infineon, Motorola called it Freescale, Philips called it NXP, and I could go on. So, that was the belief of these days, and that changed when suddenly someone at the West Coast stuffed a mobile phone into a laptop or the other way around, and you suddenly had a data display device. And you went into this on-demand world. You needed big storage, big compute, the cloud, and you needed to have these remote control devices, and you could get your Uber, your pizza, your climate control from wherever you were.
And now what we are seeing is we are entering the third step, this world that my marketing people would say the world that anticipates and automates. So, the little devices around you take over these tasks for you. So, climatize your house before you are there, adjust your insulin levels in your insulin pump when your glucose patch says, "Andre's levels are not okay." So, all of these type of topics are going now to an autonomy stage, and that is where our electronics fits in.
Interesting. So, I was looking at a report that we are putting out together with Rhombus Capital. And I just wanted to show you a couple of slides just to get your sort of what do you think about it. So, it's comparison to the US and Europe. This is the sector that that is called future of compute. And you can see although the European numbers are growing steadily, it is significantly less in terms of the actual amounts than in the US. So, like I said, almost 3x less this last year. So, is that a problem? What's your take here?
I think it makes sense to de-layer that a little bit. What is at the moment creating an enormous tailwind for the semiconductor market is big data where the ChatGPTs of this world are living. But these big data centers, they are super energy hungry, supercomputer hungry. And they are also good for the European semiconductor players because we are often supplying not the big data shufflers, the GPUs and NPUs for the AI, but we are building this control plane, the power plane, so the infrastructure. And 80% of such a data center is infrastructure cost.
So, we are participating in all of that. Nevertheless, the big digital companies from the mobile phone world, from the PC world, graphics cards world, they are sitting outside of Europe and they are benefiting from that at the moment ad-hoc trend for data centers. Now, in 10-15 years from now, I guess most of this intelligence of these 50 billion smart connected devices that we are all dreaming of, that are coming into the market, most of that intelligence is in these edge devices. And we will use the cloud, but we are not pumping for every fridge that needs to buy 2 L of milk—we are not pumping all data into the cloud and get that answer back. Energy-wise, that is not sustainable, not cost-wise.
And out of the 50 billion smart connected devices, largely 45 billion are purpose-built: so, your insulin pump, my fridge. And they don't need ChatGPT in the cloud. So, therefore, what you are doing there is you're really big time getting now tailwind behind that trend. That is at least what we believe, and this is why we are on that road.
Now, 10 years ago, we all had Chromebooks, and we had the cloud, and we were operating Microsoft Word in the cloud—you on your intelligent keyboard, I on mine. Today, no one would come to that idea. You have your laptop, I have my Surface or whatsoever, and we have Word running on our systems. We still need the cloud because you and me want to work on the same document, and we are synchronizing. I bet a similar trend is happening to the AI world.
A lot of the compute of the real reactions at the edge, in real time, physical, safe and secure, privacy is protected—so you're not pumping data around—energy-efficient. But for synchronization, how is your house doing? What is a house in the Netherlands seeing and 5 hours later I get the same weather in Hamburg? So, these type of things, of course, we are networked, but not in a religiously fundamentalist way. And this is why I think a lot of the players on the European side are going to get tailwind if this trend is materializing.
So, in your view, Europe is a bigger player in the more like low-power sort of scene.
Europe has certain key assets: safety, security architectures. The majority of today's microcontrollers is coming from Europe.
Yeah. There is players who are supplying that market already now on the world market scale. So, from that angle, that is not bad at all. The only point is you hear it already how many words I need to explain that complicated structure. For an investor, it is so super simple to say the smaller the transistor, the more trillions you get in valuation for your company and all the rest in the middle I don't understand and I forget. Compared to a 20-minute pitch that I'm doing here at the moment. And that is of course the stuff where we have to work on.
Right. But then there's also another slide and that is robotics investments. Just to come sort of a little bit deeper into this. So, robotics the situation is even more dire. Like we have a 10x difference with the US. Does that chalk up to the same thing that you just explained? And is it a problem?
