All eyes are on Nvidia as earnings are set to be reported on Wednesday, May 22, after the closing bell. Spear Invest Founder and CIO Ivana Deleveska joins Asking for a Trend to break down the chip company’s earnings expectations and how the market might react.
“We’re pretty confident that the company will beat expectations, however, what we’ve been seeing this quarter specifically is that companies beat and raise, but the market, no matter what they report, finds it pretty disappointing. So it wouldn’t be surprising to see that to be the scenario,” Deleveska says. She adds that the upside for Nvidia will come in the next two years as more data centers begin to pop up and that its chips are “positioned very well both for training and inference.”
Deleveska explains that as the AI race heats up, there are unique opportunities for companies that provide thermal management: “These are really big, massive areas of investment that are going to benefit equally from the data center investment cycle.”
For more expert insight and the latest market action, click here to watch this full episode of Asking for a Trend.
This post was written by Melanie Riehl
Video Transcript
We’re just days away from the videos.
Make a break.
Earnings report Estimates are sky high for the chip maker, implying a year over year Quin Tup of earnings per share, and analysts are expecting it will handily beat those forecasts as it has in the past five quarters.
Our next guest expects a solid beat for the company on Wednesday.
We’re joined by Spar Invest founder and chief investment officer Ivana Deska.
It’s great to see you again, Ivana.
Thanks for being here.
Thanks for hearing.
Me too.
So, um, you were looking for the company to beat again, but there I guess there is a difference between beat and a difference between beat and raise and the shares going up, considering that they’ve already what almost doubled this year thus far.
So what are you expecting on that front?
That’s right, Julie.
So we’re pretty confident that the company will beat expectations.
However, what we’ve been seeing this quarter specifically is that companies beat and raise, but the market, no matter what the report, finds it pretty disappointing.
So it wouldn’t be surprising to see that to be the scenario.
However, I would say that the upside for NVIDIA is really gonna come in the next two years.
So if you look at estimates 1 to 2 year route, they’re pretty conservative, and we think they’re gonna have to meaningfully come up.
The data centre cycle is just in the early innings, and that’s what’s driving this, uh, upside in in earnings estimates.
And we expect that they’re going to deliver closer to 40% kegger rather than the currently 20% ker that it’s in the analyst consent for the folks who don’t know kegger compound annual growth.
So in terms of the themes we’re looking for the earnings.
There’s definitely the data centre that is a huge part of it.
I think that comprises something like 80% of their revenue.
At this point, they still have a gaming business that we should be watching for as well.
And then it seems like the other thing that folks are gonna be watching for is Capex, which has been a big theme among big tech companies.
But we’ve perhaps has been a little bit less of a topic for NVIDIA.
Do you think that’s gonna be more important this quarter?
Well, Julie, there is a pretty big difference in the business model of the hyper scaler versus NVIDIA.
Right?
The hyper scaler being the cloud service providers like Amazon and A with AWS and Microsoft with azure.
Exactly.
So these companies really rely on Capex to drive their business.
So what they’re doing is they buy in video chips and they basically sell them as a service.
So they’re gonna have to really run their Capex if they wanna be growing their revenues.
For NVIDIA, it’s a little bit of a different story.
They don’t need to be investing in capital expenditure as much.
So they’re gonna be investing more in R and D and keeping up with the with the cycle.
But Capex is not as big of a driver for them just because of that dynamic.
And we’re not showing Capex that we showed operating expenses which are ticking up on a year over year basis.
That’s right.
And that would be including R and D, right?
So R and D would be, uh, would be increasing.
Um, I wanna ask you about training versus inference, which is something I hear a lot of talk about here.
So we’re in the stage of a I right now where everyone’s building their models and the A I, um, large language models are learning, which is the training phase.
Then we’ll get to the inference phase where those a IS start making predictions start making leaps of their own, if you will.
And I think one of the questions has been I mean, un, un indisputably NVIDIA has dominated the training phase.
What happens as we get further into the inference phase?
You seem pretty confident that NVIDIA is still gonna be pretty dominant here.
Well, their chips are positioned very well both for training and inference.
So we don’t really see them being disadvantaged on the inference side for inference.
You can use a wider variety of products.
So that is maybe where the where some of the confusion comes from, where people are thinking, OK, well, you do you really need to be using the highest end Blackwell chip, right?
Couldn’t you be using a competitor’s chip?
And that is true, so you can use a wider variety of chips when doing in France.
However, that market is actually growing pretty significantly, and NVIDIA disclosed that it represented 40% of their, um of their data centre.
Revenues today is being used for for in France, and this was two quarters ago.
So that number may have gone higher over the past, Uh, the past quarter.
So they’re well positioned in infer.
I don’t think it’s necessarily the case that, uh, competitors are more likely to, uh, to disrupt that market.
And a MD did highlight on their, um, Ernest call or one of the conferences that they attended, that they are actually trying to position their next generation chip to be better at model training.
So I think over time we’re gonna see similar, uh, similar performance in both from from both competitors.
Um, and then in the 30 seconds or so we have left beyond NVIDIA, I know you’re looking at sort of two main themes that will fuel outside of NVIDIA, where a I is going.
So that’s right.
So we are seeing, actually, uh, within the data centre there is inside the rack, so this would be the chips that we’ve been talking about.
But there is also opportunities outside of the rack, which is companies that are providing thermal management, whether it’s liquid cooling, companies that are providing power generation power components.
So these are really big, massive areas of investment that are gonna benefit equally from the data centre investment cycle.
And then there’s also B to B that you’re looking at, right business to business.
Kind of which is a little overlap there.
But that’s right.
So our ETF more broadly invests in B two B technology.
So within B two B technology, there are several areas.
Hardware is one area.
Cyber security is another area.
Data infrastructure is a third area.
So what we focused on here is the data centre hardware cycle.
Right?
But there are many more opportunities beyond that.
All right, thanks so much excited for Wednesday.
Thank you, Ivan.
Appreciate it.
Add Comment