So I think, Austin, you add to we've had probably the most ex Intel people on this podcast, so we're happy to have another one and just excited to get your take as someone who who was there.

Yeah, at this point, at this point.

Well, you know, I'm wearing my Intel blue today.

So let's see.

So Intel.

So early on, you know, once NVIDIA started selling all the GPUs and making a ton of money, everyone was saying, where's Intel? Where's Intel's GPU? At the time, I kind of said, hold the phone.

Intel is a CPU company.

My guess is that people want to know how to trade in the stock market and are trying to get some intel.

Is that something that you should always do? Do we always need to split up prefill and decode? Or are there just certain applications where it's really good to have speed, I'm willing to pay a premium for speed, therefore I'm willing to go through the hassle of splitting these things up and having orchestration software and having different types of silicon and all of the things that are associated with that versus just running it in the GPUs or the system that I've already bought?

Yeah, yeah.

Yeah, there's like so many different nuances here.

So if we take a look at the hyperscalers, when they're deploying, you know, tens of billions of dollars of GPUs, and as they're started, which, of course, early on GPUs, very general purpose, very flexible, you know, gives you the freedom to change your workloads.

But let's say you're an open AI, and you're like, like, no, these are our specific models that we know that we are iterating on, like the biggest, the frontier one, and the medium-sized one and the small one, you start to say, hey, we should really sort of cater to this workload's needs.

And therefore, it would make sense to deal with the complexities that you pointed out of having to split up, pre-fill, decode, orchestrate that.

might be worth it.

And so ultimately, do we end up where enterprise is actually the big market here instead of cloud?

Matt, I want to hear your thoughts.

yeah a lot of a lot of similar arguments there i mean at the end of the day meta is an ads-based business and they have always used ai compute to drive their ads business and they're very good at that so yes they're spending a lot of money on capex but that those gpus can right away go to drive their business i think in this world of llms Meta needs to own their own frontier model because it can impact how they train their ads models.

So they kind of need to own their own destiny there.

So right there is enough of an argument for them to train their own frontier models.

And then I think what investors have a hard time with is just the time to return on those investments.

monetizing that AI.

And of course, it's ultimately with their core business, either improving their ads or creating new surface areas for their ads.

Tell us about something like that, which seems impressive and lucrative to you.

What does this tell us about AMD's strategy, broadly speaking, in its next phase of products?

Hmm.

So AMD's strategy.

So AMD has always been a high-performance computing platform company.

They have a portfolio.

They have a portfolio of CPUs that they want to offer.

They have a portfolio of GPUs.

dimensions.What's the reception been from customers to this Helios rack system?

all right awesome one of the companies that's trying to slot themselves in there is arm they're shifting from like licensing ip purely to actually selling these full racks for this ai why is that important how is it meaningful to uh earnings power Yeah, totally.

In the data center that Jensen showed, Jensen and NVIDIA actually had racks of CPUs that were ARM licensed CPUs that NVIDIA makes.

Jensen's rationale was, hey, actually CPUs are becoming the bottleneck for GPUs because there's still a lot of work that needs to be done on the CPU to keep the GPUs fed.

ARM saw this as an opportunity for them to also not just license their intellectual property to people like NVIDIA for those CPUs, but to actually make silicon themselves.

And this is a big opportunity, a huge opportunity.

Historically, ARM, for the last 35 years, has only licensed their intellectual property.

And then within the last handful of years, they took it a step further.

So what did you make of some of the developments that we've seen just in the last several sessions with the memory names?

Does that kind of offset the cost at all? Is it getting something in return? H-how do you think about that?

Yeah.

Yeah.

So AMD said this is tens of billions of dollars of revenue per gigawatt.

Um, if you crunch the numbers and if you're just looking at like if you think about these warrants ultimately as sort of a discount, um, you could say, okay, well, how much of the company is Meta getting in return? How much is that worth, and what is sort of the discount off the price of the chips? Uh, on the call this morning, Lisa pushed back on that thinking, say-saying, "Hey, that's too narrow-minded." The goal isn't to extract as much revenue from Meta today as possible.

The goal is to really co-design and develop and for AMD to really win with Meta as a first big best customer.

They're also doing this with OpenAI and, and that AMD will benefit that-- from that in the long run and sort of capture that value later with all the other customers.

I mean, this is really turning into just the big getting bigger in some ways, right?

Which is the street rewarding more right now? Because we got evidence from Microsoft and Meta that suggests, I don't know, and then Apple's kind of unchanged.

Yeah, no, this is a good point.

And this is sort of a tale of three companies.

Meta, they are an AI company at the end of the day.

They sell ads against engagement.

How do you get engagement? With recommender systems that know what to put in your feed.

And then... getting the right ad to the right person at the right time is also based on AI.

How did you assess what we learned from Tesla?

Is that something then, if we take that as sort of, hey, expectations are sky high, more likely to disappoint than exceed, is that something that time can correct? Because it kind of sounds like we're basically also saying, hey, these stocks are expensive, but they're expensive because there's a lot of expectation that they're going to have to deliver on.

Yeah, there's definitely a time component here.

And there's the horizon.

We can only see so far with respect to the applications that'll take off, the consumer applications.

So there is a capability overhang where models are being trained, enterprises are adopting AI, and they're getting value out of it.

We see CapEx spend is really high, but there's not yet really a killer app or something that is going to take us up to that next side.

where consumer use might skyrocket or enterprise AI adoption might happen even faster.

How big of a story is this? Or is the story pretty straightforward now and Nvidia already made its play?

If we were explaining it to somebody who wasn't in tech,

Yeah, totally, totally.

So when GPUs have to talk over long distances, it's better to use light instead of electrons.

There's fundamental reasons why that's more efficient.

When GPUs are really close together and they talk to each other, historically they've talked using electricity over copper traces.

But now, as the industry starts to push more and more data around at faster and faster speeds for training and for inference, there actually will be some physical limits to communicating over using electricity, essentially.

And how the chips talk to each other, basically.

Do you think anything like that could be in the cards for 2026 or 2027?

Austin, what are your thoughts here? I mean, what's going to drive these vehicles going forward?

Definitely like Ivan said, you know, I think charging infrastructure is something people think a lot about.

Honestly, it's also affordability.

There's also, so, you know, can people who are looking for a $20,000, $30,000 vehicle, can they get an EV? And then there's also form factors.

My family, we have four kids.

Can we get an electric minivan that my wife loves? So I think as people get in and experience vehicles, as automakers bring prices down, as the charging infrastructure get better, I think we'll see adoption take off.

And of course, like Ivan mentioned with autonomy, I think autonomy is really coming to EV vehicles first.

What about names like Uber and Lyft or Rideshare? Austin, what are your thoughts there on that kind of autonomous driving or what it could mean, like names, those names in 2026, 2027? Yeah,

You know, Austin, as you look at this, is NVIDIA OK or is NVIDIA at risk in this, you know, continued downward movement of volatility we're seeing?

Hey, Marley.

So NVIDIA is doing great.

Like you said, earnings were a blowout.

And if we look to hyperscalers earnings calls, you'll see that the big four, Meta, Google, Amazon, Microsoft, they all said they're going to increase their 2026 capex spending.

They're accelerating.

And we're talking $100 billion for some of them in 2026.

How much of a threat are these actually to NVIDIA's current market dominance, in your opinion?

So what do you make of this recent weakness?

Great question.

You know, I think there's been some noise lately.

I think what really what triggered it is we had earnings from the hyperscalers.

We had Amazon, Microsoft, Google and Meta, of course, in the past couple of weeks where they said they are accelerating their build out to the tune of $100 billion in 2026, roughly summer, you know, 70 to 95 to even 125 billion.

of CapEx.

And so I think all of a sudden investors started to ask, okay, wow, those are really, really serious numbers.

And so as far as like recapping what we've already learned from earnings, are there any themes that stuck out to you that clearly resonated as far as like performance the next session or following, you know, several days? Or is there any themes that stuck out that maybe didn't work so well?

And the question in my mind really was, okay, he's talking about price performance as a way of evaluating how good tranium is, is tranium actually a better chip? How does it actually step stack up against Nvidia CPUs from from what you're seeing?

Yes, so you're right.

Tranium is an internally designed chip.

You can think of it similar to Google TPUs, like you said.

You can think of Anthropic as Tranium's first and best customer.

And if you're comparing the GPU to Atranium, let's call it an XPU.

That's kind of the industry vernacular.

How are you expecting the AI story to come back with Apple? Just as you're watching the company and its products, are you expecting it to come more into the conversation? Because it feels like it's kind of drifted out of focus for the company.

You know, how it sort of compares, what you're going to be looking out for as far as Alphabet is concerned with regards to what the others say as well.

Yeah.

How does Google Alphabet compare to the other Mag7? You know, I think We are definitely in this generative AI phase right now.

And you've got some people who are trying to monetize that.

Tesla on one end is looking way forward with robotics and autonomy.

On the other end, you've got Microsoft saying, how can we bring generative AI to our enterprise customers? And I think Google's in a really sweet spot where they are already using AI to bring advertisements to customers.

They've got massive distribution.

What are you expecting there?

What does that tell you about the maturity or the baseball analogy of where we are on sort of how many innings into this game we are? Good question,

Alex.

Always good to chat to you.

You always bring the tough questions.

So to your point of now there are companies who are competing with NVIDIA, so AMD and Broadcom, and when they have good news, moving in opposite directions.

So maybe they go up and NVIDIA goes down.

I think that tells you that the sentiment is that we're sort of in this zero sum, like if one goes up, one's going down.

And so I think the million dollar question is then of all of these companies, which one right now do you all like the best? Like which one's seemingly still the leader in this entire AI build out space?

Who's got the balance of advantage here, and how much of the upside for Intel is just that NVIDIA's paying attention to them?

Hey John, thanks for having me.

So you're totally right.

These systems both in the data center and client side are CPUs plus GPUs.

And notice there's two product announcements.

One was about opening the AI data center door for Intel and Nvidia saying, hey, come on in.

We wanna help you build rack scale nvidia based data centers and then likewise kind of um to help each other out is very mutually beneficial intel saying hey let me open the door and bring you guys into the the client landscape so i definitely think that it's mutually mutually beneficial it's clearly the world's largest company um signaling a big vote of confidence in intel both in their products that they're going to jointly take to market also in intel's foundry which manufactures Intel CPUs.

Your thoughts?