Nvidia, Grace Blackwell, that's all ARM. Apple's iPhone, that's ARM. Samsung's Galaxy phone, that's ARM. Pixel's Google phone, that's ARM. Microsoft shipping Surface tablets based on ARM.
ARM Holdings CEO Rene Haas at the World Economic Forum in Davos, Switzerland. His position in the AI ecosystem as the licensor of the CPUs that build data centers, iPhones, even wearable tech.
It would be hard to find a company that doesn't use ARM. His vision for innovation as one of the building blocks of AI.
Where does AI go? Does AI start to replace workers? Does AI start to invent? At what point does AI create what is the next product? Plus the potential for AI to invent new AI, new companies. Haas says it is only a matter of time.
The level of innovation in the last 10 years compared to the previous 50 or the previous 100, it will get there. The question is when.
I'm CNBC producer Cameron Costa. SquawkPod reports from Davos 2026. ARM Holdings Rene Haas begins right now.
For the past couple of years, the hot topic at the annual World Economic Forum has been AI. Joe Kernan, Becky Quick, and Andrew Ross Sorkin have spoken to many leaders and regulators in the space. Execs from Anthropic, OpenAI, Meta. Those conversations will all be on your SquawkPod feed if you're interested. This next conversation though, it's with one of AI's builders. ARM Holdings CEO Rene Haas sat down with Joe, Becky, and Andrew on our outdoor set in the snowy Swiss town of Davos to talk about the big world that he is building.
Rene, thank you for being with us today. It's really a pleasure to have you here.
I know what you all do is a little different than what so many others do because you all are in the licensing business. But I saw a statistic that blew me away. Something like four chips for every human on the planet was sold by ARM in the last year. Is that accurate? Am I getting that statistic correct?
Yeah, something like that. There's some crazy statistics. It says enough ARM chips have been shipped to cover every person who's ever lived on the planet, and then some. What we do is a CPU. It's a compute platform that's the heart of just about every device you can think of, whether it's smartphones, data centers, automobiles, washing machines. Our business model is licensing. So, we license that CPU intellectual property to someone who makes a chip. That chip goes into the device. A lot of devices end up shipping.
So, how should we think about you in this really competitive chip space at this point? Who's your partner? Who are you competing against? How does that work?
It's ironic that we're in Switzerland. ARM is sometimes referred to as the Switzerland of the semiconductor industry. It would be hard, Becky, to find a company that doesn't use ARM. Nvidia, Grace Blackwell, that's all ARM. Apple's iPhone, that's ARM. Samsung's Galaxy phone, that's ARM. Pixel's Google phone, that's ARM. Microsoft shipping Surface tablets based on ARM. So, we supply our CPU IP to everybody in the industry. And the reason that that is important is that a CPU is the digital brain of every modern device. Everything needs a CPU. So as a result, ARM is pretty much everywhere.
And what do you think when you sit back and look at these wars? Do you care how this competitive landscape plays out? Do you just figure it doesn't matter? You're going to be there no matter what?
To some extent. I think right now it's an amazing time in our industry with everything going on. Obviously with artificial intelligence, it's hard not to talk about it. We see artificial intelligence workloads running everywhere, whether it's inside a data center or in your smartwatch, and that's a huge opportunity for ARM because as more complex AI gets, energy efficiency is everything, which we're really good at and that's an area we're quite focused on.
So what is your latest advancement in that exact area in energy efficiency?
Yeah. So we do the low-powered processor that's the heart of the Grace Blackwell platform. And Jensen then also talked about Vera Rubin at CES a couple weeks ago. That's all ARM, that uses six times the number of ARM CPUs in the platform that the previous platform did. But we're also in small devices like wearables that consume less than a watt of power. And I think over time it's going to be very critical to be able to run those AI workloads in small devices such that we're not consuming so much energy in the cloud.
The rush that's on right now for any of the data centers that are trying to make sure they have enough electricity, that they have enough copper, that they have all of these different ingredients, does that impact you at all? Because it does impact your customers.
Sure. Yeah. Anybody who's in the semiconductor supply chain is impacted by this. Whether it's not enough memory, not enough wafers, not enough turbines, to some extent, we're all impacted by it. But what's happening is that because the amount of compute that's required is just so large, that's why you're still seeing pretty large numbers being reported by all the chip guys just because the volumes are so darn big.
Andy Jassy told us this morning that the whole idea of the shortage for energy still exists for data centers, but it's better than it was 18 months ago. Would you agree with that?
I would agree that it's better than it was 18 months ago, but one of the problems that we have particularly in the US is not so much energy because in the US we actually have a fair bit of energy, but it's actually getting that energy to the places where the data centers need to be. Transmission lines that require contracts and easements and all the things that our industry never had to worry about. Some small village having to get an easement to get a new transformer, cut across county lines. So yes, some things have gotten better, but still the scaling problem exists.
