Yahoo Finance

AMD today closed on a $4.9 billion deal to buy AI infrastructure company ZT systems in a move to bolster its data center and AI offerings. Joining me now for an exclusive here on Yahoo Finance, that is AMD CEO Dr. Lisa Su. Lisa, always nice to get some time with you here. So, you had announced this acquisition in August, closing today. How does this change AMD's business moving forward?

Good morning, Brian. It's great to be here with you guys. And today's a fantastic day for AMD. We are super excited to be announcing the close of our acquisition of ZT systems. Um, ZT is a phenomenal, phenomenal team. And we are now, when you think about, you know, AI infrastructure, you know, what we've learned is that, you know, customers really need, uh, really an end-to-end, you know, AI capability. You know, hardware, software, systems, really to get AI, you know, in practice as soon as possible. And you know, ZT is a key part of our overall toolbox. And um, it's super great to have it closed today.

You're making a key distinction. Um, Lisa, you're saying you may have strategic partners soon for the manufacturing business. Now, you've made it very clear, at least to me, and I think investors, you don't want to be out there making your own chips, which is, I think stands in stark stark contrast to what we're hearing from Intel trying to figure out what it's trying to do on that front. Why don't you want to make your own chips?

Well, at the end of the day, Brian, when we look at, you know, the ecosystem, the ecosystem is, you know, we want to be best at what we do. And we are best at being a design company, at being a collaboration partner, at really defining reference designs and capabilities. You know, AI, there is so much need for computing. I think we see that all over the place. You can see all of these new applications coming in. We're designing best of breed, the highest performing chips out there that are going into the largest supercomputers. And we partner really, really well. We partner in silicon manufacturing with ZT systems. We're bringing in an incredible team. We're bringing in over a thousand design engineers who are working hand-in-hand with our silicon engineers as well as our customers, uh, systems engineers so that we can put together these world-class systems. And yes, we are going to look for a partner for the manufacturing capability because we really do believe manufacturing is best when you can go across a number of different vendors and not just one. Now, from a ZT manufacturing standpoint, we're super excited about the world-class asset we have here. It is US manufacturing at its best. It's AI systems at its best. And so overall, it's just great to be able to bring such a talented team into AMD.

How long do you anticipate, uh, for that, I guess, that review process to take? And when do you expect the deal to happen?

So, the way to think about it is, uh, today we're closing the overall ZT Systems acquisition. So it's coming into AMD. Uh, we've been very active with strategic partners. Uh, there's a tremendous amount of interest to partner with our ZT manufacturing capability because of its US manufacturing base. And you'll hear much more about that, um, you know, in the coming months as we get through that process. But the main thing is, you know, we now have the ability to offer AI systems and AI solutions sort of end-to-end. And that's exactly what we want to do when we look at our next generation road maps. We're getting ready to announce our MI 350 series. That's coming up soon. Next year, we have the MI 400 series. We're going to even broader capability. So, lots of AI coming from AMD.

When I saw this deal originally, I thought this was Lisa's and AMD's play on AI infrastructure being continuing to get built out at a rapid clip. Since last August, we've heard a lot of concern about things slowing. Are you concerned that we're seeing, or are starting to see a slowdown in AI infrastructure buildout?

You know, Brian, I want to be really clear about this. Uh, there is no question. We are in the very early innings of AI. Now, you know, put some of the noise aside. What we continue to see is more applications, more capability. Enterprises are just at the very early innings of adoption. And frankly, they need more help, and they need more help from folks like ourselves. They need more help on the software application layer. They need more help to really bring sort of the capabilities so that you can see business value. So I am a big believer in this is the very, very early part of AI. Uh, the need for compute continues to be immense. Uh, we see that throughout all of our customers globally. And we're going to continue to invest strongly in this area because I think this is the single most important technology. I like to say it's the single most important technology of the last 50 years. You know, certainly what I've seen in my career is I've never seen a technology go as fast in terms of both innovation, um, innovation on every level of, you know, hardware, software, models, applications, as well as when you think about adoption, we're still so early in the adoption cycle, and there's so much more we can do.

Why, why do you think those fears are out there though, Lisa, that things are starting to slow down? Are you seeing that, I guess within your MI 350 orders or your current AI chips at all?

Look, I think there is um, a lot of learning that's happening in the market. You know, people are trying to understand, hey, what does it mean when the new model comes in? What does it mean when deep sea comes in? What does it mean when the next chat GPT comes in? And I would say that they're probably a little too short-sighted in the conversation. You know, the long arc of this AI wave is enterprises can do much more, can become much more productive. Uh, this is not just about how do we save some labor costs? I mean, I think that's the uh, the small piece of the pie or the large piece is how do we see true business model innovation? How do we build better products? Um, how do we have, you know, how do we have the ability to solve problems that frankly now we can't do with traditional computing that we can now do with AI plus all of the general purpose compute. So I view this as, you know, people are just learning as to what the value of the technology is. And you know, we're going to see a lot more benefit as it gets more broadly deployed.

This interview, Lisa, comes at a very interesting time for the world. Lots of focus on tariffs. Now you get almost 65% plus of your business is international. You contract with TSMC. Help investors understand what's at stake here for the semiconductor industry on the tariff front.

Well, overall, if you look at the baseline, the baseline is, you know, we do want a more resilient supply chain. You know, resilient supply chains are important for every industry. They're especially important for the semiconductor industry. We are big supporters of US manufacturing. I think having more US manufacturing is a positive thing. You know, obviously when you're running a global company, we're going to be watching the, you know, what happens in the tariff situation very closely. And the key is to be agile, to have a very broad supply chain. And, you know, we are going to work hard at bringing more manufacturing back to the US as well.

Are investors wrong to think tariffs equal higher prices for semiconductors? And then by extension, PCs are going to cost more, mobile phones are going to cost more?

I do think that there could be some short-term impacts as you think about, you know, what happens in the overall tariff situation. But I don't, you know, I think it's too early to say what the longer term impacts are. Uh, I think we have to look at how things play out over the next, um, number of months. And, you know, the key is, as I said, we run a global business. Uh, you know, we ship all over the world, and we need to continue to have supply chains that comprehend that as we go forward.

Does, does your, do your, will your chips cost more?

From what we see today of the current tariffs, obviously we have to see what happens later in the week. Um, I would say there's a relatively smaller impact. Um, we'll have to see what happens as we go through the rest of the week and what happens as we go through the next number of months.

Lisa, um, this is really, I would say, your second really large acquisition. Of course, you go back to Xilinx in 2022. I believe that was $35 billion. What have these deals helped you do? And then what might be next on your radar?

Brian, I have to say I'm really proud of our team from a couple of angles. First of all, organic growth is really our mantra. Like we are a growth company, and from a technology standpoint, that's about putting out great products, ensuring that we keep a very strong roadmap. But we also have an incredibly strong balance sheet. And what that's allowed us to do is make very strategic acquisitions. Um, very, very happy with our Xilinx acquisition if I think about what that has added to us in terms of customer set, product capability, skills, you know, 5,000 engineers came into AMD. Now with ZT Systems closing today, another, you know, 1,000 extremely skilled engineers coming into AMD. Systems capability, the ability to really give us end-to-end AI. But we've also made a number of other acquisitions in networking, in software, bringing in talent. And it's very important for a company of our size to have the ability to, you know, one, grow so quickly organically, but also layer in additional businesses and additional skills as we go forward. And this is all about ensuring that we are the leader in high performance and adaptive computing going forward. There's so much need for compute, and we want, you know, AMD has the ability to provide compute across all of the different areas. And I like to say in computing, there is no one size fits all. You know, you need great CPUs, great GPUs, great FPGAs, great networking capability, great ASIC capability, and the ability to mix and match that. And that is our strategy.

Lastly, Lisa, my contact I talked to on the ZT deal, uh, they told me that this will allow AMD, this deal, uh, to leapfrog Nvidia in a few years. Are they totally wrong?

Well, we certainly are looking at how we drive the best in high-performance computing systems. And like I said, Brian, this is all about ensuring that we have all of the pieces, both from a hardware standpoint, lots and lots of investment in software. And ZT is a critical piece in that supply chain and that value chain to give us the systems knowledge. And when we're working with our customers, our customers are super excited at doing co-development with us and our ZT team as we think about what are we going to do in the future together.

All right, we're going to leave it there. Dr. Lisa Su, AMD CEO, always nice to get some time with you. We'll talk to you soon.

Thank you, Brian.