Among many of the signals that the data center buildout is here to stay. While AI compute is becoming a tradable asset class. And coming off of earnings season, we saw Nvidia's data center revenue spike 117% year over year. We're also seeing Qualcomm reach a $60 billion deal with Amazon to supply compute, as well as infrastructure for Amazon Web Services expansion.
And after the clues today, investors will be turning to Oracle's earnings to gauge if its free cash flow can keep up with its heavy AI CapEx plans. While joining me to expand on where the AI buildout stands and what the market impact is, is Johnny Wu, founder and CEO of xETFs. Johnny, good morning. Thank you so much for joining us.
While earnings season is beginning to wind down and we saw a huge data center, revenues as well as guidance capture the expected acceleration of AI sales for many tech giants out there. But where do you see the AI build out heading and what does this mean for equities?
Sure. Well, first of all, Remi, thank you for having me. Um, pleasure and an honor. Um, so our view, um, broadly, is that we're still early days in this AI data center. Build out this AI revolution. I compare it to a little over 25 years ago when the internet was being built out. Right. And you had a lot of companies, um, investing in laying down the fiber and the cables to connect the world.
Right. And back then, you know, some of the larger names that were really important were names like Nortel Networks, Cisco Systems. Right. And one of the biggest beneficiaries of the internet. And clearly a winner was was Amazon. Right. Um, it's it's it's good to see now today. Right. There's 2020 hindsight.
But of those three companies Amazon did amazing right. Nortel networks went bankrupt. Right. And Cisco for many years just traded sideways. And then more recently it's picked up. But it hasn't outperformed the S&P 500 as an example, right. Today I see the AI data center build out. There is going to be a continued mass spending CapEx spending by all the hyperscalers.
You're seeing that in their quarterly reports, right. And you're beginning to see profits now from some of the beneficiaries, right. So namely Nvidia, but you also see downstream from that all the other picks and shovels, companies that are going to continue to benefit from all this massive spending by the hyperscalers.
Right. To the tune of, you know, more recently, if you if you take the top five hyperscalers in the 600 to 700 billion range, projected to go over $1 trillion over the next couple of years. So we think that the spending will continue to happen. We think that we're still early, right, in the AI build out. But clearly you're going to have both winners and losers.
And um, yeah, that's the market environment today.
Yeah. And Johnny indeed, there are so many analogies out there comparing past tech revolutions to what we're seeing when it comes to the artificial intelligence infrastructure built out. But I do want to get your perspective on emerging markets. So how has South Korea become a key part of the global AI semi supply chain?
Yeah.
Sure. Well, to build a data center right. Obviously you need to have the land right. You need to have the people to actually build out the data center. And then you need all the individual parts and the power and the energy on the part side. Um, you know, what was critical was you need processors, GPUs that Nvidia makes, that AMD makes and others make to actually do all the compute.
But the compute itself, um, stands idle unless it has, um, memory, uh, to be able to, to, to actually access data. And for AI accelerators that are needed in these AI data centers. Um, you need high bandwidth memory. So HBM. Right. So that theme has been around right for the last year, but really front and center for the last 5 to 6 months.
And there's only three players globally that dominate HCM. Right. And the leader is SK Hynix in South Korea right. Samsung Electronics also headquartered out of South Korea and in the US. It's a micron technology right. So South Korea because of its history where they invested in semiconductors long ago, right over 30 years ago.
They just have a significant advantage over, um, you know, a lot of other countries in terms of semiconductors and memory and specifically HBM memory. So it's been one of the better performing stock markets, um, geographically over the last year, year to date. Lots of volatility as well too. Right? So, um, it's not for the faint of heart, right?
But clearly South Korea is a leader in memory and will be critical going forward as well.
Yeah. And Johnny, you mentioned a key word there. And that is volatility. But when it comes to investing in artificial intelligence I do want to get your perspective on ETFs. So there are many funds out there. So in about 60s. Can you tell us what specific funds you are watching right now.
Sure. Um I think ETFs are an amazing wrapper, and it's clear to become the investment vehicle of choice for many investors and a lot of direct retail investors. Right. Um, you know, it's it's often said that there are more ETFs than stocks now in the US. Right. Over 5000. And that is true. Right. What ETFs allow you to do is express a specific view or a tilt.
Right. If you don't follow the markets if you don't do your homework right, Warren Buffett says the best investment you can probably make in equities is just to buy a broad based passive index like the S&P 500. And I agree with him. right? Um, but if you do have a specific view, right. Just take technology.
Right. Technology over the last ten, 20, 30 years has been the clear winner. Right. And an index like the Nasdaq 100 has outperformed the S&P 500 over that time period. Right now it comes with volatility though. Right. And and clearly that's it. So on the on the on the passive end you can express it. But if you want to go more specific like express views and on AI thematics there are a number of ETFs that allow you to take more specific whether that's in in specific industry like semiconductors, or you can go more refined.
Well, Johnny, I am sorry to interrupt you, but we are all out of time for today. So thank you so much for joining us. I appreciate all of your insights as well as your perspective.
