ARM's CEO Discusses Podcasting, AI Opportunities, and Industry Challenges

ARM is poised for immense growth as AI integrates into every aspect of technology.

### ARM's Strategic Evolution and Opportunities in AI and Chip Manufacturing
* 🎙 The CEO of ARM recently launched a podcast called **Tech Unheard**, aiming to engage with industry leaders and share insights about the tech landscape, with his first guest being **Jensen Huang**.
* 💼 ARM was acquired by SoftBank for approximately **$32 billion**, and its current market valuation stands at around **$160 billion**, following a successful IPO just over a year ago.
* 🔄 The company's business model has evolved from a **horizontal licensing approach** to a more **vertical strategy**, focusing on specialized products for sectors like **data centers and automotive**, which has enabled ARM to **tap into new markets**.
* 🤖 With **70% of the world's population** interacting with ARM technology, the CEO believes that AI presents a **massive growth opportunity**, as all AI workloads will ultimately run on **ARM architecture**.
* 📈 Despite competition from Nvidia, ARM is already integrated into advanced data centers, and the CEO envisions ARM's role in AI extending from **training algorithms to inference** across various devices.
* 🌍 The CEO expressed concerns about the **concentration of chip manufacturing**, particularly with TSMC's dominance, advocating for a **more diversified and geographically distributed semiconductor industry**.
* 💡 He highlighted the importance of government involvement in chip manufacturing investments, particularly in light of the **enormous capital expenditures required for new fabs**, which can reach **up to $35 billion annually** for leading companies.
* ⚖ As ARM navigates legal challenges with Qualcomm, the CEO emphasized the complexity of competition within the industry, stating that ARM's focus remains on **delivering solutions that meet market demands effectively**.

That's why I have you. And thank you all for joining us. We have a lot of questions for you, so we'll go ahead and get started. First, I begin. You are not just the CEO of aam; you're now a podcaster. All right, you began in your spare time, I guess. You began a new podcast called Tech Unheard, and your first interview was with Jensen Huang. That's right, which I gave a listen to. First of all, what gave you the idea to start a podcast? And secondly, how do you find the time as CEO to do something like that too? Yeah. So first off, thank you again for having me here. I'm an amateur at this, so you don't have to worry about your day job as far as podcasting goes. You know, we were talking with some of our board members about how to use a unique medium to talk to leaders, talk about the armed story. And one of the ideas that was suggested

was a podcast. I was a radio disc jockey, actually spinning records in university. So, I had a little bit of experience and aspirations to do that kind of work early in my life. So, in talking with some folks, you know, around the team, we said, you know, let's try this and let's do something a little bit more innovative and talk to other leaders and give them a perspective of maybe a conversation that they may not naturally have with someone who's professional like you are. So, yeah, we kicked it off. Please listen, it's on Apple Music, Spotify, and Jensen was my first guest. Great, great. Congratulations. And, you know, in terms of the business arm, was acquired by SoftBank for about $32 billion. At some point, SoftBank agreed to sell it for about $40 billion. That deal didn't go through, of course, in video was supposed to buy it.

That deal didn't go through. Now, you went public just over a year ago, companies worth roughly $160 billion, which seems like a successful IPO. But could you talk to us about what the business model for ARM has been and how it's evolved over time, over that period of time? And especially, what's the strategy going forward? Yeah, well, thank you for the you know, for the nice words. I'm a 30-year-old plus company. Our business model for most of our time on earth, if you will, was very horizontal in that we had a licensing model. We develop IP. Our products are primarily CPUs. The business model is very horizontal in nature, meaning that we designed a very general-purpose product, very power-efficient product, and then license it to customers independent of the vertical market. And in other words, we didn't really target a certain vertical. I took over the core IP business right after SoftBank bought ARM, and we pivoted

to a more of a vertical approach, meaning that we now have products very specific for the data center, very fitting for automotive, very big variety. Smartphones were born. And that diversification really allowed us to expand into many new markets with brand new products. SoftBank buying ARM in 2016 allowed us to invest in those new areas. So fast forward the result. What you're seeing now in terms of performance is that strategy. You know, going forward as we think about the next number of years, what we're seeing in the marketplace is a need for more solutions. Chips are getting increasingly more difficult to build. It takes a long time to design a chip, takes even longer to fabricate the chip. So the more that arm can do to help customers get to market faster is helpful. So we're now developing solutions, what we call successes that get people to market sooner and allow them to get more profitably in a sooner way.

Mm hmm. Okay. I want to talk about A.I. and, in particular, the opportunities in AI. There is some debate, as we've referred to, I think, in a couple of these sessions, about whether there's a bubble. And I'm going to assume that you think that there are real opportunities there. But can you talk about what ARM's role is there and how you see the company evolving to take advantage of those opportunities? So, one of the most unique things about our company is the fact that the device that we build, the CPU. We estimate that 70% of the world's population touches ARM in some way. We're in security cameras, automobiles, PCs, smartphones, data centers. And what that means is every workload, whether it's an application, an operating system, or A.I., runs through ARM in some way. So, that means A.I. is going to run on ARM, period. And we're the only company on the planet

that can run those A.I. workloads from the smallest devices on the edge. Again, a wearable glasses watch phone to the data center. So for us, A.I. is a gigantic opportunity in terms of growth going forward because AI is going to be everywhere. And as far as a bubble goes, I just don't ascribe to that. I look at things such as when people say, Gosh, is it overhyped? Or the stocks have gone down last quarter. That's almost like saying, Should I short AT&T in 2000 because the Internet is not going to happen? Or if Ford was a stock in 1907, would I short them? Because the automobile is not going to happen. It's just not. I mean, I have seen such advancement across the air in the last couple of years that the amount of innovation that it's going to drive is just going to be incredible. Okay. Okay.

The chip company that probably has taken the most opportunity has been able to capitalize the most on AI so far as in video chip maker in your space. They now have a near monopoly in terms of the highest-end AI accelerators. How do you see that evolving in the future? And are there opportunities for ARM in particular to be able to play in that space and drive real revenue from? Yeah. So they're already invidious most advanced chip called the GB 200, which is Grace Blackwell. Microsoft just announced that they're going to be deploying Grace Blackwell, the most advanced Nvidia chip in their data centers that uses the ARM CPU. So Grace is the ARM CPU. BLACKWELL is the Nvidia GPU. So ARM is there already. So we are going to play in the most advanced data centers using an AI ARM. ARM will be there. I think what's most exciting for us, as I mentioned though, is the fact that not only will we be in the data center for

training these algorithms. Ultimately, the training has to be transferred into inference, and training is the teacher. Inference is the student. Inference is actually running the workload and inference actually running that AI agent that's going to happen in the data center. That's going to happen in the car. That's going to happen on your watch. That's going to happen in glasses. That's going to happen in phones and everywhere it happens. It's going to run on arm. So it's going to be a gigantic opportunity for us. The data center is just one place. Mm hmm. Okay. So you see arm having opportunities kind of from the beginning to the end, more than often, more than opportunities. The agents will run locally. And the reason for that is you'll have a hybrid situation where some of that will run in the cloud. Some of it will run in your local device. Again, your glasses, your wearable, your phone.

But locally, what you'll be able to add is security, things that can make sure it's private to you, that not all your data is being passed to the cloud and vice versa. So yeah, for ALM, I think it's going to be just an amazingly large opportunity. Okay. As we were talking about before, I just moved from Japan, where we spent a lot of time looking at SoftBank and especially Masayoshi Son's ambitions. We've written stories about how SoftBank is looking at ways to enter the chip market and particularly to compete within video quite directly in GPUs. He has a secret plan, his analogy that we've written about in the past. How does ARM, isn't still 90% owned by SoftBank? How does ARM fit into that equation? Yeah, so no product launch announcements today. So nothing I can say about products coming out that you've written about. I can say that on the SoftBank board member, I do advisory work for SoftBank.

I talked to Masa all the time. It's no secret that he is a big believer in AI, has been for quite some time. He's talked about it, actually, probably as long as any of the folks out there. I think his vision, in terms of where it's all coming together now, you can start to see the pieces of it, as I just described. Where does aam fit? All those A.I. workloads are going to run on aam somewhere, somehow. So, that's the reason that we spend a lot of time talking to SoftBank about the future. Okay. Okay. Earlier, we had the ASML CEO on, and he was talking about some of the challenges that they're facing in the market. They surprised investors last week, as we talked about. And it does seem that there's a little bit of a disconnect in the market, where many people see opportunities. ASML reported that their orders were about half what analysts had thought they were going to be.

Seems there are certain areas of the chip industry that are struggling quite a bit. And on the other end, you have Sam Altman talking about how we need more capacity in the chip industry. We need billions of dollars, maybe trillions of dollars of investment. What is the disconnect, and how does that get resolved? Yeah, so there's a lot to unpack there. Where ARM plays, as I mentioned, because 70% of the world's population uses ARM, and I think 300 billion chips have shipped since we started that have ARM inside. And every year, about 30 billion chips. We have a unique view into the industry because just about every digital device uses ARM. There are some markets that aren't accelerating as fast as others, and candidly, those are the ones that aren't using A.I. in a very large way. But in areas that can use A.I. and will use A.I., people can't go fast enough.

And I'll give you an example. Many of the hardware devices that are being introduced today were designed two or three years ago. That means the chips were designed two or three years ago. The chips that are going in your phones or your wearables. Those were all invented before CBT was released or before Lamo was released. So what we're seeing on one end of the industry is just a rush to develop more and more chips that have this air capability. Now it takes time. You know, chips aren't built overnight. In fact, it takes from the time we engage a partner, you know, two or three years at least until that comes to market. So I think you'll see some of these new products coming out. I think what Sam is talking about is just generally the ceiling to get to AGI. Artificial general intelligence requires more compute. And the more compute you have, the better the models get. So I see. I see both.

But generally speaking, I'm pretty optimistic about the growth for our industry. Okay. Okay. We were talking about this a bit before. One of the gating factors is making sure that you have chip manufacturers who can make the most advanced chips. TSMC seems to be gaining market share right now as a couple of the key competitors struggle, Intel in particular. And Samsung also has been struggling a bit. They're both trying to move into the foundry business a bit in competition with TSMC. What are the dangers around having TSMC garner so much market share in foundry? And is it good for the industry, or would it be better if there's a more diversified manufacturing base for the industry? Yeah, TSMC is an amazing company. They've just done so much for the industry relative to the innovation, how they work with partners, etc. etc. If you remove semiconductor manufacturing or foundry for a moment

and just step back and say, is it healthy for any industry to have one dominant supplier and then add on top of that a dominant supplier that's largely concentrated in one part of the world? Not, not really. So as a result, I think it's necessary just in terms of diversity of supply base, to have more and more advanced manufacturing and geographic diversity. Fabs. The United States. Fabs in Europe. Fabs in India. Fabs in Saudi Arabia. I think over time, we're going to see not only more suppliers, but geographic diversity as well. Okay. Governments obviously are investing a lot in that very question. You're seeing the United States do it. You're seeing Europe do it. Japan has spent a lot of money on it, too. Do you think that's making progress at this point? Do you see light at the end of that tunnel? It's the capital expenditures

required for these fabs are enormous. TSMC, Intel, Samsung, I think their CapEx is anywhere between 30 to $35 billion a year in new factories. And if you start looking at the revenues that these companies produce, you start very quickly looking at, oh my gosh, that is a lot of money to invest. Now, on one hand, you can look at it and say, well, the demand will be there, etc., etc. But it doesn't take much in terms of a year slowdown in demand and the companies have a really negative financial outcome. So, yes, governments have been involved, I think should be involved in my opinion. We saw in the U.S. for the very first time a policy act around CHIPS Act to put funding into the fabs, which I generally a big supporter of. I think it's necessary. Okay. Let's talk a little bit more about Intel and their struggles. You grew up in a chip industry where

Intel was really the dominant player, kind of set the pace for technology evolution over time. Now we're openly talking about whether Intel is going to be broken up, maybe sold into pieces. What kind of opportunities are there for ARM within that space as Intel goes through these struggles? Yes, we work very closely with Intel. It's a bit of strange bedfellows. Many folks would consider ARM and Intel to be natural competitors on the product space. But what's very important to us, back to this discussion around supply base, is that Intel Foundry Services (IFRS) is successful. We do a lot of work with Intel as we do with Samsung and TSMC to ensure that our products can be manufactured there. We work very closely with their engineers to ensure that the leading-edge technology can run on ARM. So we want to see Intel successful, and we want to see if it's successful. I think it might be strange to hear the ARM CEO say that, but we want Intel

manufacturing to be successful. Okay. Bloomberg reported that ARM took a pretty close look at buying a part of Intel. Would you like to comment? Nothing I can say on that one. Okay. Okay. All right. Thank you. Thank you. I wanted to turn. You talked about moving up the value stack, kind of moving beyond its traditional place in the industry. And as you have moved up the value chain, you've gotten closer to doing what some of your customers do. Right? Getting closer to being able to do these complete designs. And. Yes. It speeds up innovation, as you were alluding to, but it also puts you in closer competition with some of your customers. You have a lawsuit with Qualcomm in particular? Yeah. Isn't it in both companies' interest to settle that and to resolve the legal dispute at this point? Or what would you say? These are tough questions. It's a good thing I'm a podcaster, so I know what's coming.

Right. You know, first thing on, competing with our customers, you know, it's rather complicated because if you look at some of our customers, our customers are Amazon. Our customers are Microsoft. Our customers are Apple. Our customers are Tesla. They all build chips using ARM. I'm not going to build an electric car. I'm not going to build a phone. I'm not going to build a data center. So to look at the value chain relative to who builds chips, relative to whether your end business is a chip business or a product business. It's gotten a lot more gray. We follow what the industry is demanding, and what the industry wants to see is solutions getting to market faster. And that's what we're focused on. Qualcomm. Not much I can say on that, other than we're headed to a trial. I think it's the third week in December. We feel very good about our case.

We think our case is quite simple and straightforward. And as I said, December, we'll find out. Okay. So you head to the courtroom then. One last question. I think we have time for this. In this audience in particular, there's a lot of interest in whether ARM, which is historically a British company, whether they would ever look at doing some sort of dual listing. Of course, we've heard about this many times in the past. A year ago when you did your IPO, you decided you would do it in the U.S. instead. One of the opportunities for listing here, what are the pluses and minuses of that sort of thing? Yeah, we did. We listed in New York over a year ago. And at that time our comments were we would look at a secondary listing. We're still open to it. It's not something that's top of mind right now, quite candidly.

But we'll continue to have discussions with both stakeholders in the government and in the local exchange. We'd love to find a way at some point in time, but also at the same time, there's a lot of other things that keep me busy relative to my day job. Okay. All right. Rene Haas, thank you very much. Please join me in welcoming. Thank you. Thank you.

Arm CEO on Intel, Chips, AI, Listing in US

Welcome to the Bloomberg Technology YouTube channel. Subscribe for original series and mini-docs on game-changing technologies, profiles on innovators, live conversations with experts, and breaking technology news & analysis from Bloomberg TV.

Bloomberg Technology

The CEO's Journey: From ARM to Podcasting

The CEO of ARM started a podcast called Tech Unheard to share unique conversations with leaders, drawing on their past experience as a radio disc jockey.

ARM's business model evolved from a horizontal licensing approach to a more vertical strategy, targeting specific markets like data centers and automotive.

The company's success is now focused on providing solutions to help customers build chips faster and more profitably.

ARM's CPUs are integral to 70% of the world's devices, making the company essential for running AI workloads from edge devices to data centers.

The company's unique position allows it to capitalize on the growing demand for AI, with opportunities for growth in training and inference across various devices.

ARM is already involved in advanced data centers through partnerships, and the hybrid model of AI workload distribution between the cloud and local devices presents significant revenue potential.

ARM, 90% owned by SoftBank, plays a crucial role in the chip industry, with 300 billion chips shipped since its inception.

The company's unique position gives it insight into the industry's dynamics, as 70% of the world's population uses ARM.

While some markets are struggling, those leveraging AI are experiencing rapid growth and increasing demand for chip capacity.

The Chip Industry's Evolution and Challenges

The hardware devices being introduced today were designed 2-3 years ago, with a similar timeline for the chips used in phones and wearables.

TSMC is gaining market share in chip manufacturing, raising concerns about the risks of having a dominant supplier concentrated in one region.

There is a push for geographic diversity in advanced manufacturing, with governments investing in chip production across the United States, Europe, India, and Saudi Arabia.

The Enormous Investment in Chip Manufacturing

Leading chip companies like TSMC, Intel, and Samsung are investing $30 to $35 billion annually in new factories.

The demand for chips is crucial, and a slowdown can lead to negative financial outcomes for these companies.

ARM is working closely with Intel, Samsung, and TSMC to ensure successful manufacturing and is also moving up the value chain, which has led to competition with customers like Qualcomm.

ARM's Complicated Customer Landscape and Future Listings

ARM's customers include major companies like Amazon, Microsoft, and Apple, making their business model complex.

The company is focused on delivering solutions to market faster, in response to industry demands.

While ARM is open to a secondary listing in the UK, it's not a top priority at the moment, as the CEO is focused on other business matters.