Lisa Su explains what's coming next in AI

[00:01] AMD at its Advancing AI conference in San Francisco. The unveiling of new chips, a new $5 billion investment in Anthropic, and a reiteration that demand for AI remains incredibly strong. I sat down for a conversation with AMD chair

[00:15] and CEO Dr. Lisa Su. Su has been the driving force behind AMD's ascent the past decade. >> First of all, for the industry, I mean, we're calling the industry market growth to be, you know, get to a $2 trillion

[00:28] um as we think about, you know, from now through 2030. So, it's just an incredibly large market. Um we're seeing this point in AI, you know, we've called the fact that uh AI is at an inflection point where more and more people are

[00:42] doing useful work, doing meaningful work, and you know, that has made um inference uh sort of the largest growth driver overall. And I am super excited to say that we are in full production for Helios, our

[00:55] MI455 rack-scale architecture, our new Venice uh CPUs, and the tremendous amount of new content. So, yeah, it's it's a pretty exciting time for us. I industry, um and more importantly for all of us in the AI ecosystem, like this

[01:10] is the time to really um accelerate what we can bring to the market. made uh at the event, Lisa, uh this one also stood out to me. AI is evolving, you talked about just now, too. AI is evolving from training to inference.

[01:25] revolution, and what does it mean for a company like AMD? >> Well, what it really means, Brian, is all of us are now starting to experience the real power of AI. So, uh we have great customers that are training and

[01:40] doing foundational models, um but the key is, you know, the model is only as good as the output that you get from it. And so, with inference, that is all of us asking questions, that's agents working, and that is growing at a um

[01:54] really, really accelerated rate and pace. And that's why we're saying that just the accelerator market alone, it's going to reach $1.4 trillion by the time we get to 2030. And inference is really the way we turn, you know, AI from a

[02:10] technology to something that really changes the way we do business, the way we do research, the way we do healthcare, all of those aspects of it. So, I'm really, really excited to see

[02:23] the rate and pace of adoption of AI. >> Lisa, I think I've told you this before, vicariously through your X feed. I see you traveling the country. I've seen you stuff things you've been doing this year. As you've traveled the world and

[02:37] you've done your important work at AMD, what has surprised you about this AI movement? >> I I think what has, you know, surprised >> I I think what has, you know, surprised me, Brian, in a very positive way is the

[02:49] rate and pace of adoption is actually faster than any of us thought. I mean, about what's going to happen in a market. We think about updating it on an market. We think about updating it on an annual basis. And with AI and with

[03:03] world with our customers, what we're seeing is things are changing on a monthly and a quarterly basis. So, just over the last five or six months, you know, agents have become really, really productive. Like the idea that you're

[03:16] not just asking a AI model a question, you're actually asking an AI model to help you solve a problem, help you, you know, code you know, a new software module or you know, help you plan your next trip.

[03:32] year ago we wouldn't have thought were possible. And all of that requires more and more compute infrastructure. So, that is probably the most interesting, exciting, surprising thing is for those of us in the industry, you

[03:48] know, I've never seen a technology adoption curve like what we're seeing with AI right now, and agents are just taking it to a whole new level. >> You really you increase your total addressable total addressable markets or

[04:00] TAM for your data center business, your accelerator market are pretty significantly as you look out to 2030. On the topic of AI agents, in 2030, what are these agents doing that they aren't doing today?

[04:14] >> Well, let me start with the very important thing to remember about AI is, you know, AI is a tool. So, AI is a tool to make every one of us more productive, to make every one of our companies more productive.

[04:28] think about the productivity you can get from 10 employees. I mean, you got really really smart people. But if each of those 10 employees had, you know, 10 agents or 100 agents or even 1,000 agents, they can be much much more

[04:43] productive. And so, you know, the whole idea of agentic AI is to be able to create sort of a closed-loop situation where you can ask your AI to

[04:55] actually help you solve more complicated problems so that, you know, you as an employee can just be much more efficient. Like, at the end, you know, the human is the one that decides what is the right or wrong answer, but AI can

[05:07] help you look at a whole set of options much much faster and more efficiently than you have than if you had to do it by hand. And that's the power of agentic >> Within the the the technology that you unveiled at this event, Helios for

[05:20] example, you just you just mentioned it. What what does something like that do products don't do? >> Yeah. So, one of the most important things when we think about AI is just how many questions can you answer and

[05:34] how much does it cost? So, you know, you hear things like AI tokens or, you know, dollar? You know, what Helios does is it's an incredible feat of engineering that allows you to scale much, much uh

[05:49] broader in terms of the size of the models, the types of questions, you know, the number of users that can use a certain system. And so, the way to think about it is we're seeing more than a 30 times improvement in performance when we

[06:04] look at the new Helio systems. And, you know, from a customer standpoint, that just means you can get a lot more done, you know, for your money. And, the way I think about it is it's a very positive feedback loop, right? The more

[06:18] AI can do, the more compute you need, the more compute you have, the more intelligence you have, the more problems that you can solve. And all of that relates to the importance and the essential foundation is the computing

[06:31] layer, which is what we do at AMD. >> You You've mentioned this to me in the your keynote. You And it's one of your philosophies or one of your beliefs that one chip player will not win it all

[06:43] for this AI during this AI revolution. Why is that still the case? >> Well, I think the most important thing to think about, Brian, is like the world is a very heterogeneous place, right? We're all different, our companies are

[06:57] different, our problem sets are different, what we're trying to solve is different. And as a result, there's not just one chip that can do it all or one company that can do it all. What you really need is you need an ecosystem.

[07:10] You need an end-to-end set of engines, so that's what I like to say at AMD. We have an end-to-end set of compute engines from CPUs to GPUs to FPGAs to ASICs. The ability to put all that together. And then the other thing that

[07:24] we have is we have a true belief in an open ecosystem. So, this is the idea that the more developers we have on AMD, the faster we can move the ecosystem, the more progress that we can make. And, you know, that's just good for the

[07:37] big believer in there's no one-size-fits-all, but there's an incredible power in bringing the ecosystem together and allowing us to innovate simultaneously.

[07:50] >> You mentioned too as well that quote customer demand for these new technologies is extremely strong. And I look at how from a stock market perspective, we just saw folks react to higher CapEx estimates from a

[08:03] and an Alphabet. What are investors missing? Because I talked to you, I all tell me demand is strong. So why are we so concerned about hyperscalers >> Yeah, I have a true belief on this. We have to

[08:19] look at the long arc. You know, the long arc of technology is I am absolutely convinced of the power of compute. And you know, when I say I think AI compute equates to intelligence and why wouldn't you want more

[08:34] intelligence? Of course you want more intelligence. Now, the important thing is where is the return on investment for those investments? We are seeing the return on investments. I mean the fact is we are ramping our

[08:46] own AI usage within AMD very significantly, you know, month over month and we're seeing the productivity come back in just better products, more capable products, faster time to market and that will come

[08:59] out as you look at the long arc. Now whether you see that on a quarter by quarter basis, you know, obviously you know, that's that's for others to think about, but from my standpoint, the long arc, it is absolutely clear that compute

[09:12] demand is at a premium today. It is one of the places where you can't just wake up tomorrow and say you want more compute. You have to make those investments, you know, 12, 18, 24 months

[09:26] in advance. You have to plan the entire supply chain and the entire ecosystem for that. So we are very confident in the demand picture being there and we are ensuring that we have all of the pieces to ensure that um

[09:39] you know, we can satisfy a larger and larger piece of the just brought up. So you're not seeing there's no leveling out. You're not seeing anything slow down. >> We are not seeing that at all Brian. I

[09:54] mean every conversation I have with every large customer is how can we go faster? And I I like I said I think it's a fundamental belief in we believe in the tech and it will return on investment

[10:06] cycle. >> So you also this week too getting lost in the sauce a little bit is this new deal with Anthropic. You're investing $5 billion into this company. Now you had you gave warrants to to Meta, you gave

[10:18] warrants to OpenAI. Why that that change here with Anthropic? Why invest $5 billion in them? >> Well first of all Brian I have to say I'm thrilled with our new partnership with Anthropic. I mean there's no

[10:31] question that they are you know one of the leaders in frontier AI models and being able to work hand in hand with their engineers is you know really a wonderful thing.

[10:44] The fact is you know MI250 Helios is a fantastic product. We're excited to be ramping with them at scale up to 2 gigawatts and think about it as every relationship is different right? We believe in Anthropic. We are very happy

[10:59] to be making a strategic investment of up to 5 billion as we work together and ecosystem is we're all in this together. So it's hardware, software, models, bringing it together in data centers, ensuring that

[11:16] we are really tracking those things together and and so that's exactly what we're doing with Anthropic. It's really aligning our strategic interests with their interests and you know to your broader question

[11:28] you know we're thrilled with our overall partnerships across the board. When you look at open AI when you look at meta these are the top AI companies in the world that are choosing AMD yes because of the technology but also

[11:42] because they believe in our long-term roadmap and the ability to collaborate and work together to solve the biggest problems in AI compute going forward. >> Is there any concern that everyone in tech is so intertwined?

[11:58] right? I mean I think it's an ecosystem where we're working on this together and in some sense I think we are uh probably more aligned which is actually a really good thing when I say aligned it's because these

[12:14] are such large investments over time the more we know about what our customers need the better prepared we can be for for that moment. prepared we can be for for that moment. So I think that's what the the AI world

[12:28] is such that you need every part of the ecosystem to come together and and so the fact that the ecosystem is intertwined is the way it should be and efficient long-term. >> Lisa lastly you said something on your X

[12:43] really resonated with me and I think ties into everything you announced that your parents came to this country course of your career I mean when you got this job I remember when you

[12:56] the meta deal open AI deal now in the topic now this moment with Helios these moments in the tech industry have you had any do you reflect on on how far

[13:08] had any do you reflect on on how far you've come? I view this as the most exciting time of my entire career and I'm extremely view it as a true privilege and honor

[13:25] compute right now Um, I love the fact that we can contribute and make a difference. I love the fact that the technology that we're working on is impacting billions and billions of people every day. And um, I couldn't ask

[13:37] >> Well, uh, keep uh, keep rocking there, Lisa. It's always great to get some time I'm sure you're busy. You're you're the star of the hour. I'll talk to you soon. >> Thank you so much, Brian. >> Take care.