WHEN: Today, Thursday, September 14, 2023
WHERE: CNBC’s “Squawk on the Street”
Following is the unofficial transcript of a CNBC exclusive interview with SoftBank Founder & CEO Masayoshi Son and Arm CEO Rene Haas on “Squawk on the Street” today, Thursday, September 14 to discuss expectations for Arm’s IPO – the biggest IPO of 2023 & more. Following is a link to video on CNBC.com: https://www.cnbc.com/video/2023/09/14/arm-ceo-china-business-reflects-growth-we-see-in-the-rest-of-the-world.html.
PART I
RENE HAAS: Yeah so, Arm of 2023 returning to the public markets, we’re a very different company than we were in 2016 when we were largely associated, as you said, with mobile. And the important thing to think about Arm or remember is that we were born from building a device that was going to run off of battery. So that sensibility about power efficiency is in the DNA of our engineers. So fast forward to 2023, when you look at the diversification of our markets, cloud data center, automotive, everything with EV and automotive. These require extremely power-efficient processors or CPUs, which is what we do. So whether it’s data center, obviously around sustainability, you want to have as much low power as possible. Again, back on these cars running off batteries, it’s a great place for Arm to really grow our business. And that’s what we did in the years being private between 2016 and now. We diversified our business. We’ve got significant growth in the cloud data center and in automotive. And then with AI, AI runs on Arm. It’s hard to find an AI device today that isn’t Arm-based. Google Alexa – excuse me, Amazon Alexa for example, that device which does voice recognition, etc., that’s AI. And what we see happening going forward is products that didn’t have a CPU to run AI, they’ll need AI. You might need more CPUs to run more complex AI, so we see just huge growth opportunities there.
DAVID FABER: What is running generative AI though, is more GPU and is obviously all produced by Nvidia. There is a relationship there obviously, you once worked for the company as well. But investors I’ve spoken to would like to see a lot deeper relationship. Is it your expectation that you’re going to sell more to Nvidia? Because they’re not that large a customer.
HAAS: So Nvidia today is doing obviously a lot around AI – generative AI and training. Their latest product that they announced the Grace Hopper Superchip, which is their accelerated GPU for AI training, now uses 72 Arm CPUs as the core CPU on that. Remember, no type of accelerated training for AI can run without a CPU. So now the combination of the Arm CPUs with Nvidia GPUs we think will be even more growth opportunity for us going forward.
FABER: But you know, when investors looked at your S-1 for example and or your recent numbers, they didn’t see a great deal of growth year-over-year. And yet you’re pointing to a much more significant growth next year over this year and even after that. Why?
HAAS: The year-on-year growth was not that great, as you said. If we look backwards for the last three years as we’ve pivoted our strategy, we’re about 15% year-on-year growth going forward – excuse me, for the last three years. But going forward again, when you look at these mega trends of efficiency, software, complex AI all having to run off batteries or low power devices, this is huge growth for us. And again, David, Grace Hopper is a great example of the kinds of devices that can only be built on Arm.
FABER: Masa, it wasn’t that many years ago that you were selling this company to Nvidia or at least had plans to sell this company to Nvidia. That obviously did not happen. The price tag then was 40 billion. A lot of it was made up of Nvidia stock. What has changed between then and now?
MASAYOSHI SON: We did not want to really sell. It was the Covid that made me really, you know, go into the protective mode. So I had to go protective mode and I had to choose the more conservative, careful operation of SoftBank. So that we were selling to Nvidia, but the deal was actually one-third cash, two-thirds was exchange of Nvidia’s share as a combined company of Nvidia and Arm. I believed in the future of Nvidia back then, and it was right. And I believed in the combinations of the power of the two companies would be enormous. I believed in the future of AI. And it’s really now getting proved. And this is the beginning of big AI time and Arm is going to have a big role in that.
PART II
HAAS: Our China business reflects the growth we see in the rest of the world. We’re seeing huge growth in the data center around cloud computing, also with AI, and then EVs. Huge growth in China in terms of EVs. And China wants a lot of what the rest of the world needs: power efficiency, software ecosystem. A lot of the same software that’s used across the world is used in China. So what we’re seeing David, in terms of our China market growing, is largely around those two areas: data center and automotive. Now in terms of the broader issues, I think I share the same headaches that just about every tech CEO does these days. We comply, of course, with all the regulations that come down relative to export control, if there’s something that we need to adhere to, of course. But it’s really it’s a tricky market to figure out just in general because of all the things that are going on geopolitically. But broadly speaking, our China business has been doing very well.
FABER: Yeah, I want to get back to the business itself. But Masa, you know, your – Softbank owns a significant stake in the joint venture that I’m describing in China. You’ve obviously done business there. You were a very large holder of Alibaba for many years. What is your sense in terms of the risk that China poses given the percentage of revenues that it comprises for your company for Arm is quite high?
SON: U.S., China is having a very complicated situation now. The, China has significant impact to the economy of the rest of the world. So I think – I hope the situation get better, but who knows. I just am one of citizen who is wondering and, you know, concerned about the future of China, U.S. and the rest of the world.
FABER: Yeah. What do you hear? I mean, you have had relationships, deep relationships with many people in the business community there. How deep are your concerns? Not just in terms of the back and forth between the U.S. and China, but also in terms of the regime itself and its crackdown, so to speak, on entrepreneurship, if I could call it that.
SON: Well, it’s difficult to comment. Whatever I comment, it goes into all kinds of headlines here and there, so I have to be careful what I say. But our exposure in China, Softbank as a group has reduced significantly. Because now, you know, most of the shares in Alibaba from Softbank is already sold.
PART III
SON: Yeah. We – SoftBank owned 75% of Arm until just a few weeks ago. And just a few weeks ago, we bought back from 75% to 100% from Vision Fund. We pay even higher price. So I’m more confident about the future. I think the value is gonna have a good upside, really long term. And that’s why I bought back with a higher price than $51. And I just want to have all the investors have a good time going forward.
PART IV
SON: Well, Rene has, you know, a great chemistry with me. We share the vision completely together. We’re talking almost every day, we’re chatting with WhatsApp and, you know, calling all the time almost really every day multiple times, like three times, five times, sometimes 10 times a day.
PART V
SON: Yeah, well, I’m a big believer of AI. Since I started SoftBank, I was a big believer of, you know, the microprocessors enabling all kinds of technology, evolution starting with PC, and then mobile, internet, and then now going into AI. It is a front end of all the innovation of our industry. It’s the core is the microprocessor. So Arm is now going to become the core of the AI revolution. My belief of the front end of the IT industry’s evolution is really shifting to AI so my my focus, my belief is all centered in AI and Arm is going to become the core of that. Many of the Vision Fund portfolio companies, we have about 500 companies, they’re going to be, you know, having a lot of great applications of AI and they’re going to be beneficiary of this Arm and many other AI technology players.
FABER: Yeah, well, I mean you recently said or not that long ago, Arm is positioned to benefit from AI becoming the, quote, “The basis for a new society.” What does that mean?
SON: Yes. Well, I think this is the first time that mankind experience something smarter than mankind itself. Right? Mankind was the smartest animal on the earth. The AI is going to surpass and surpass big time. I think the AGI stage is coming very soon. And once it comes, it goes so far away. And it’s a new stage of society, that we all have to ask ourselves, what is mankind? What is job? What is life? What is the intelligence? It’s a new experience completely new society that we are going to face.
FABER: Masa, if you’d been in that meeting with a lot of the leaders in AI that just took place with Senator Schumer and other members of Congress, what would you have shared? I mean, you just said it’s going to exceed our own intelligence. There are those who are quite frightened by that prospect and worried about the dangers of AI. Do you share that?
SON: Well, if we mishandle it, it has a danger like automotive society. It has a danger of the car accident if you, if you don’t regulate. The automotive society is regulated with a traffic light and speed limit and don’t drive with alcohol drinking. So AI society should regulate to protect humankind. However, it has more merit than the demerits. So I think I’m a believer. I’m optimistic that AI is going to solve the issues that mankind couldn’t solve in the past, like difficult disease, the natural disasters, the car accidents, all kinds of other issues that humankind had in the past will be helped by the advancement of AI and technology.
FABER: Rene, you’re going to be in the middle of this. I mean, you know, you’re going to be providing designs that are going to help Nvidia use, create chips that are going to be powering generative AI. Do you feel similarly or are you perhaps more concerned?
HAAS: I agree with Masa from the standpoint that we are going to see some very, very profound advancements around AI going forward. The ChatGPT moment, if you will, I think was a tipping point relative to what the capability of these large language models could do. So on one level as a CEO of a company that builds a lot of devices based on AI and an engineer at heart, phenomenal opportunity. And I think for us going forward back to the CPU being the center of everything and you can’t really run AI without a CPU, it’s going to be a huge growth opportunity for Arm. That said, there are a lot of social and ethical type of things to consider, which I think as a society, we have to figure out and we’re in early days on that.
PART VI
HAAS: I think what’s important to remember again, our core product is the CPU and what is the superpower of the CPU? It’s the broad software ecosystem around it. We have a software ecosystem like no architecture probably ever invented. Every major operating system whether it’s iOS, Android, Windows, Android Linux, Automotive Linux, all the Linux distributions for the cloud, not only do they run on Arm, but they’ve been optimized for years, in some cases decades. RISC-V, as you mentioned, is a new emerging entity. Right now, what we see is that they just don’t have much traction in the area around software ecosystems. It’s not to say that we don’t watch our competition very seriously. We obviously do and there’s a lot of war room discussions about how to address that. But broadly speaking, I think the the hurdle they have to cross is around the software ecosystem. And that is not an easy moat to cross.
FABER: Why not?
HAAS: Just the sheer amount of development time that goes into it. I’ll give you an example. Windows running on Arm PCs. I worked on them personally in my previous life back 2009. Here it is 2023, 14 years later, and we’re just now starting to see growth of Arm on Windows PCs. It just takes time. There’s a lot of software work that needs to be done, development time, and optimizations. And, and that’s just for PCs, when you think about phones, when you think about the cloud and you think about automotive, the software gets more and more complex. So, I think over time, that is something that we’ll continue to invest in and I think it’s a it’s a big path RISC-V to cross.
FABER: Speaking of over time, Masa, you know again, you at SoftBank own 90% of Arm, the now public company. Are you going to sell any of that? It is your expectation that you will be a holder from here on of that 90% stake or is it something that you do see as potentially monetizable over time?
SON: Well, I want to keep it as much as possible as long as possible. I’m a long-term believer and and so on. Of course, you know, if the next Covid comes, who knows? But our intent is to hold as much as possible as long as possible. I wanted to keep 100% of Arm on the reason we are having IPO and selling is because Arm is such an important company for the industry, I wanted to have investors opportunity to participate on the upside opportunity of arm and wanted it to become a public, used to be a public company. So back into public company’s position so that you know people will have more transparency of the operation of the companies and so on and also give the engineers and employees the stock options to realize.
FABER: Right. Rene, you’re now a public company CEO. I mean, you do have one very large shareholder, but does it change how you do your job?
HAAS: Well, you know, as Masa said, he’s a long-term believer in Arm. So making sure I keep my number one shareholder happy is very important. But at the same time, he and I are very aligned on the on the long-term vision and view for the company. I believe we are one of the most foundational companies inside our industry. The world runs on Arm, it’s hard to find an end device that doesn’t run on Arm. And at the same time, I think when we looked at the percentage of the population that uses Arm, it’s probably north of 70%. So while we’re thrilled to be public today and I’m thrilled for all the employees as I said and our partners and developers, I really think that the best is ahead and I’m super excited about the next five to 10 years for the company.
FABER: Is there any scenario in which you would actually start to not just design chips but design your own chips?
HAAS: I don’t want to foreshadow what our product plans might be. But the company is foundational to the electronics industry and we have a lot of opportunity for growth.
FABER: That’s where you’re going to leave it, just that? A lot of opportunity for growth? We’re going to be able to do interviews in the future together. So I guess I’ll be—
HAAS: You can ask me next time.
FABER: I’ll be able to ask you that again. Let me come back to something as well Rene that a number of investors did at least mention to me as a potential concern, which is this lawsuit against Qualcomm. You know, never great to be suing somebody that you do business with. Can you give us an update in terms of your expectations and, and, and why perhaps people should not be concerned?
HAAS: Qualcomm is a great partner. You know, back to the foundational aspect of what we do, just about every product that Qualcomm puts into the market runs on Arm, their Snapdragon that goes into phones is based on Arm, their forays into PC will be based on Arm, their automotive platforms are all based on Arm so they’re a huge partner for us. We hope that dispute can be settled. I can’t say much about the litigation, it’s headed for courts next year. But I’ll just kind of leave it at that.
