Investors uncertain of big capex spending on AI: Analyst

Stifel applied technology analyst Ruben Roy joins Josh Lipton and Alexandra Canal on Market Domination to discuss the status of the chip sector amid recent volatility as Nvidia (NVDA) tries to hold onto its dominant position.

Roy tells Yahoo Finance that investors are worried about when Big Tech companies will see the payoff of their AI investments. Another key chip investor concern is the sustainability of the increased capital expenditures (capex) spending to invest into AI and where the spending will shift after data centers are built.

Nvidia stands out as a major benefactor of the AI boom. Despite recent volatility, the chipmaker's stock has more than doubled since the start of 2024.

The analyst tells the Market Domination team that Nvidia’s “checks have been solid” and “it's not just Blackwell” as demand for the company's Hopper GPUs is also holding strong.

For more expert insight and the latest market action, click here to watch this full episode of Market Domination.

This post was written by Naomi Buchanan.

Video transcript

Big tech companies are investing millions on artificial intelligence.

But it's still unclear when investors will see the return on that investment and chip stocks have been taken along on a wild ride as a result.

For more, we're bringing in Ruben Roy, applied technology analyst at ST and Ruben, we know these big tech giants, these hyper scalers are aggressively spending on A I, but we've also seen the volatile swings these past few months.

So from an investor perspective, what are the biggest concerns that you're hearing from your clients?

Hi, thank you for having me.

Uh Certainly I would say the number one question we're getting is this investor debate around, as you said, all of this investment that we've seen now going on uh I guess two years relative to ro I or returns on the infrastructure investment.

Where's the, you know, where's the beef or where's the money on, on some of the uh infrastructure that's being built out?

So uh from an investor perspective, it certainly seems like there are concerns as we look ahead to 2025 as to the sustainability of the Capex spending from the big cloud service providers, whether or not investment spending will extend into Fortune 500 enterprise and what's next for the big suppliers of hardware into the investment cycle like NVIDIA, for instance.

And so certainly that's top of mind for investors.

And uh you know, it's been reflected in the volatility, as you mentioned in the stock, NVIDIA under performing the S and P 500 over the last three months, but certainly having a very strong year to date.

Nonetheless, your top of mind for, for those NVIDIA investors, Ruben also is what demand looks like for that, that New Blackwell chip.

What are your checks suggesting there the checks have been solid and it's not just Blackwell.

There are a lot of concerns heading into Nvidia's quarterly report late August around timing of Blackwell, whether or not NVIDIA is having some technical difficulties with Blackwell.

Uh But certainly coming out of that report, certainly, I I would think would ease some investor concerns.

The numbers were solid with a, with a nice beat for the July quarter and a nice raise relative to consensus expectations for the October quarter and Blackwell hasn't started shipping yet.

I mean, we're talking still about a lot of demand for the existing hopper architecture and if you kind of look through and speak to the supply chain or look at some of the some of the supply chain commentary, certainly it would sound like demand continues to far outpace supply.

And so when we look at 2025 I think something that investors should take into consideration is that blackball is not one product, it's not a GP U.

It's a lot of different configurations of systems that NVIDIA is going to be selling into a lot of different types of A I infrastructure builds.

I would say one thing that really becoming evident, we saw open A I and build their 01 LLM over the weekend.

Uh is that the size of the A I cluster is moving higher?

And I think that that's really, really interesting if you think about uh Nvidia's positioning in A I, not just like I said, you can use the system software and um uh and, and the overall infrastructure that's required for A I.

So Blackwell demands strong, but I do wanna go back to a point you made earlier that investors are concerned about the sustainability of this spending from these big tech players.

Are there any signs that investors can look out for a ahead of NVIDIA earnings when you're looking at that demand picture and whether there is a slowdown of spending, you know, certainly in this type of spending and, and the rate at which spending has been growing, which we've never seen in covering the space for in excess of 20 years.

Uh you know, it, it's, it's not sustainable for, you know, forever, right?

So at some point, there will be digestion periods ahead, you know, what are some signs that investors can look for or, you know, signs of progress for the A I infrastructure vendors like an NVIDIA is that in, in our view, uh you know, we're, we're at A, we haven't even really broached the uh I, I think the extent to which A I is gonna impact a lot of different industries.

We're building out these uh training clusters and we're, we're putting into place large language models.

But I think there's a long tail investment cycle coming beyond this, which is going to take the data that's being generated by the language models and then actually doing something with that data.

So the so called training of models moving through inference of those models and putting that data to work.

And I think NVIDIA and others, we cover A MD which we like very much are building infrastructure around their hardware.

So adding a lot of different types of software um resources to their hardware to get ready for that longer tailed investment cycle.

And so while the numbers may not, um you know, continue to grow at the rates that we've seen.

It's quite interesting to us to see where the spend is going to flow, right?

Is it gonna flow into some areas of software that NVIDIA is expanding into?

Is it gonna flow into other areas of hardware that we haven't seen yet?

We'd like to think about A I at the edge A I moving into things like personal computers, uh mobile handset, even automobiles.

And so I think, you know, we're gonna see a lot of these companies expand their approach to A I to address a lot of these markets and benefit as the spend moves around from, you know, again, what we're seeing today, which is the core infrastructure around large language models.