Nvidia Q1 Earnings Results: Top Takeaways

Finance Frontier
2 followers
1 Video View·May 21, 2026
Chapters
00:00 Nvidia beats earnings expectations again
01:15 Data center revenue nearly doubles
02:16 China restrictions and AI chip demand
03:02 Why Nvidia changed its reporting structure
04:59 Jensen Huang on AI factories
07:32 Why AI infrastructure spending keeps accelerating
10:05 Nvidia’s next growth phase explained
12:47 Nvidia raises its dividend
AI Generated·May contain errors
It's so crazy, as somebody who's followed this company for 25 years, nobody used to know who they were.
And now you've got everybody and their brother thinking about it.
We're increasing our quarterly dividends from 1 cent to 25 cents.
Let's get to Yale Finances. Dan Howley on the latest. Dan, what are you seeing?
Yeah, Brooke, they obviously beat on the top and bottom line and they provided a better-than-anticipated Q2 outlook.
But the stock initially fell around 2%. Right now, it's just hovering around 1%.
In the quarter, for Q2, they said they anticipate between $89.1 and $92.8 billion in revenue.
Wall Street initially was looking for $87.3 billion. That's for Q2.
For Q1, though, the company came in with EPS of $1.87 in revenue of $81.62 billion.
The estimate was for $1.77 and $79.18 billion. So, obviously, a big beat there.
And then on the data center side, obviously, the most important business for NVIDIA, they had revenue at $75.2 billion.
The projection was for $73.47 billion.
Just to give you, kind of, put that into perspective, the company brought in $39.11 billion in the same quarter just last year on the data center side of things.
Now, CFO Colette Kress said that they didn't see any hopper product revenue out of China during that quarter.
You'll recall that Jensen Wang, the CEO, was just in China with President Trump trying to get more chips or at least ship any chips into the country.
The U.S. has offered NVIDIA licenses to ship, but China has to then allow them to be imported.
And so far, we're not seeing that. The reports indicate that that could change in the near future.
It's been, kind of, a touch-and-go situation for them there.
And I just want to point out that Kress also said that about 50% of the revenue in the data center continue to come from hyperscalers.
Those are the Microsofts, Googles, Amazons, Metas of the world.
The other 50% came from a variety of sources.
That includes AI clouds, industrial enterprise, and sovereign customers.
Sovereign AI, where countries buy up AI chips for their own needs, is something that Jensen Wang has been talking about a lot as being another step for the company going forward.
So, obviously, a big beat here.
You know, they're dealing with a lot of competition.
Cerebrus, rather, just had their IPO last week with their own AI chip.
We have NVIDIA with their own AI chips.
And then, obviously, Amazon and Google continuing to push further into the space.
We're seeing, kind of, a different way of NVIDIA reporting out this quarter.
They broke down revenue by market platform.
We had data center and edge computing that's different than maybe what they did before.
What does this implicate to you?
Yeah, I mean, it seems as though it's a way for them to really stand out when it comes to the data center business.
You know, previously, they had data center, and then, you know, they kind of broke that down with networking and compute.
This way, you could get a better understanding of what their networking business was.
Which, by the way, they sell their networking capabilities to other companies, even if they're not using NVIDIA chips, necessarily, when it comes to the GPU or CPU.
But that allowed you to kind of get a good sense of what was going on overall.
Now, they're breaking it down where the data center is between hyperscalers and then public cloud.
So, they call that ACIE.
And those will include things like AI clouds, industrial, and enterprise.
And then, on the flip side, they have edge computing.
And so, this is where you used to have things like, you know, visual games, things like that.
It looks as though, I mean, to me, that they're kind of minimizing where they came from with gaming.
Now, this kind of edge computing side is going to include what they say is processing devices for agentic and physical AI, including PCs, game consoles, workstations, AI-ran base stations, robotics, and automotives.
Kind of like just mushing everything together that isn't data center.
So, it kind of shows that, you know, their huge focus is on data center.
Obviously, that's, you know, where it's going to be.
That's the kind of company they are.
But it does seem as though it's minimizing that kind of base where they came from as far as games go.
You know, not a huge part of the business.
Brought in, I think, $10 or so billion in this past quarter.
Sorry, $6.4 billion on that edge computing side, up 10%.
But, you know, you have to wonder what the loyalties will look like for them on that, you know, gaming front.
Obviously, it's not the biggest part of the story.
But it is something to keep in mind, considering that they are still a huge player in that space.
And to see the evolution of this company, now seeing gaming where they really sort of began teamed up right next to agentic AI.
All sort of under that umbrella goes to show where the company came from, where it's going.
As you noted, edge computing up 29% year over year.
Data center revenue up 92% year over year.
Quite significant there.
Now, as you noted, data center includes hyperscalers and AI clouds, industrial and enterprise.
But there's so many more implications that many people are wondering what's going to be said in the call.
Specifically about that Vera Rubin chip or if there's anything that we're going to hear about Jensen Wang's recent trip to China.
What do you what do you think?
I mean, I think that's going to be the big question, you know, where they see kind of the China story going.
Right. I mean, they have in their previous commentary, CFO commentary, Colette Kress said, look, we have licenses to ship to China.
It's just a matter of China wanting to import.
And then you had this kind of trip where he was there with President Trump and Tim Cook and Elon Musk.
And, you know, the result of that seemed to be at least as far as what Trump said was, yeah, you know, they're going to focus on their own chips.
That's kind of what they're interested in at this point.
You're like, OK, well, what does that mean?
Will we never see Nvidia ship there again?
And then it seemed as though a few days later that we had gotten some reports that maybe we'll start to see some Nvidia chips shipped over there.
Jensen, prior to earnings, said that they essentially have zero market share in China at this point.
And, you know, there's these two side conversations where it's, you know, the people who want to ship into China say, look,
that'll benefit America because then you have an economic and military rival that's dependent on our chips.
It'll, you know, improve revenue overall.
And then you have the folks who say, well, why would we give them the best of the best technology that we have if it's going to leak into their military infrastructure?
And, you know, it's it's an economic rival at the same time.
So that's been been the back and forth conversation.
The folks who say we should ship it then go on to say, OK, well, they're going to just build their own chips anyway.
So why shouldn't we just benefit for it?
So it's, you know, difficult.
Dan, we're also talking about very specific kind of chip.
It's not all chips available to China.
It's the H200.
Yes.
So this is their older Hopper chip.
So they have the Hopper generation.
Then they had the Blackwell generation and the Blackwell Ultra.
So you had the Grace Blackwell 200.
Then you have the Grace Blackwell 300.
That's that GP300 that everybody talks about now.
And then we'll have the Vera Rubins coming out, I believe, towards the end of this year.
And then past that, we have Feynman, which is their next next generation GPU.
So, you know, they have this yearly cadence.
They continue to stick by it.
We'll have to see what the next one is.
But, you know, as of right now, we're still waiting for Vera Rubin to hit.
The world is going to have billions of agents.
Not today.
I mean, we're going to grow into it, but we'll have billions of agents.
And those billions of agents will all use tools.
And those tools can be like, you know, like PCs, just like us humans using PCs today.
In the future, you'll have an agent using PC.
As far as kind of the reaction to the numbers, I mean, listen,
this is a name that had performed pretty well over the last couple of weeks,
not as well as maybe the rest of the chip industry.
But that also should be expected, just given the size of this company.
But I think most importantly for, you know, the numbers here is the fact that,
hey, listen, everything remains intact as far as this AI infrastructure build is concerned.
And, you know, we're kind of running slightly ahead of expected.
Angela, that's right.
Heading into this report over the past month, we did see Nvidia stock up about 10 percent.
Ivana, I want to get to you because the forecast is pretty much in line with expectations,
but also well above what initially was anticipated in the prior outlook that the company had provided.
We're now seeing about $91 billion in revenue expected for fiscal 2027.
What do you make of this report?
What do you make of the guidance that the company put out?
Well, the guidance is stronger than expectations.
However, people usually the buy side expects ahead of that, right?
So this is why you're not seeing the stock react significantly.
To the upside, to Angela's point, I think the report is very solid.
It proves that everything is on track.
We are going to be looking to hear more about Vera Rubin and specifically for us,
Vera itself, the CPU, is going to be a very important driver.
If that can add several billion dollars of revenue in the second half,
that would be an incremental positive to the street expectations.
So those are some of the things that we're really going to be focused on learning on the call.
As I was starting to talk about at that Dell event,
Dell and NVIDIA together announced they had achieved over 5,000 enterprises
that had purchased GPU-equipped servers.
So to me, that's a sign that this technology,
which they started this process almost three years ago, is really starting to take hold.
So again, we're going to see this AI infrastructure spread
because really, realistically, companies are going to want to have access to their own compute
because there's concerns about constraints in terms of power
as well as compute capacity from the hyperscalers.
Even though they're building like crazy,
they still don't have the capacity they'd like to have and need.
So companies are going to start purchasing some of this stuff on their own.
So I think that's a big deal.
The CPU point is also a big deal.
As we move to a Gentic AI, one of the key points that Jensen and others have made
is that we're moving from what used to be 40 GPUs to one CPU down to close to one-to-one.
So that's a huge increase in the number of CPUs that potentially are going to be sold
for systems that are designed to do a Gentic AI.
And that's where all the action is these days.
And I think that's going to continue to drive this story forward.
I want to point out this from the release
because Jensen Wang did say that a Gentic AI has arrived.
It's doing productive work, generating real value,
scaling rapidly across companies and industries.
He also said this. He said that the build-out of AI factories
is the largest infrastructure expansion in human history,
and it is accelerating at extraordinary speed.
Angelo, in a note prior to this report,
you had said that you believe that Nvidia's increasing content growth story
within future data centers is underappreciated by investors.
What do you make of Jensen's comments?
And do you think, Angelo, that it's underappreciated right now still?
This is no longer a GPU company, right?
It is the fact that they are an infrastructure company,
and it's the content growth that they're going to get outside of GPUs in these servers.
Right now, it'll be CPUs.
I think you'll hear a lot about that on this call.
If Jensen's got to roll out all CPU-based servers, that's what he'll do.
And, you know, clearly, when you look at the configurations
that these Nvidia-based servers have,
the most popular ones are the ones with the Nvidia-based CPUs attached to them.
As you migrate later this year,
you're clearly going to see Grok get embedded within this ecosystem, right?
I mean, as you go into the third calendar quarter,
you're going to look at Grok, potentially,
and those LPUs into more of these configurations.
And that's going to also provide some really nice upside
in terms of the content growth story in these AI-based servers.
And then, of course, as you go into 2027, you're going to have Rubin Ultra.
And then, obviously, thereafter, there's going to be Feynman.
And if you look at each of those configurations,
there's going to be a lot more content,
semiconductor content, and even potentially,
you know, obviously, more upside potential
in terms of a pricing perspective for these future servers.
So as you continue to see an explosion of these tokens,
which, you know, I would think everybody anticipates will continue to take place,
that continues to bode favorably
if you kind of believe that Nvidia will continue to be that token king
and offer the lowest TCO relative to its competitors.
I mean, I think it's all going to come down to how quickly they can ramp the revenue,
both from Vera on its own and Vera Rubin.
So that's really how they're going to drive the revenue going forward.
I would say, though, that for investors here,
Nvidia is more of a compounder.
So as you see in the numbers,
they're able to beat and raise consistently by a few percent.
And that's how the stock is going to be up 20-plus percent for the year.
It's not going to be up 200% for the year, right?
So if you're looking at the CPU market for some of these companies,
the numbers are going incrementally higher by 100%.
So I think some of the investors here need to kind of understand
the risk-reward here is very different
because you are taking a lot less risk,
but you are getting compounding returns.
Bob, I do want to get your take on this,
because within this release,
they also did say that they increased their quarterly cash dividend
from about 0.01 per share to 25 cents per share.
What's your reaction to that?
The company is generating so much cash.
It just makes sense to be able to do that.
It makes it more attractive.
I mean, Nvidia now is a household name,
and it's so crazy.
As somebody who's followed this company for 25 years,
nobody used to know who they were.
And now you've got everybody and their brother thinking about it.
So you could argue, hey, that some of this is, in a way,
helps highlight the attractiveness of the stock for traditional buyers.
But obviously, there are economic benefits as well to it.
But to me, I think it's just a way for them to smartly leverage
some of that amazing amount of cash that they're generating.
Yeah, I mean, listen, I think this is a stock,
from a valuation perspective,
trading only about 19 times our estimate for the next calendar year.
Numbers are going to continue to go up here.
And when you look at what the free cash flow potential here of this company,
they're going to generate north of $400 billion of free cash flow
over the next eight quarters, right, over the next two calendar years.
So they're generating a ton of free cash flow.
They're going to probably be able to return a lot more of that back to shareholders,
especially as they potentially are doing less in terms of investments
and maybe other companies out there.
So we'll see in terms of, you know,
how much of that they return here as we continue to look ahead.
But I think that continues to be a big part of the story for Nvidia
as we look ahead and the growth rates start to slow down for this company,
as it will inevitably do, right?
So, you know, to us, we think, you know,
this continues to be a fairly enticing opportunity for Nvidia, for investors out there.
I think really on this call,
Jensen's really got to continue to focus on the sustainability of that revenue trajectory
outside of just, you know, through 2027.
You've got to look past that and really be able to tell that story out there to investors.
And that will hopefully continue to keep the stock going and working favorably.
Yeah, so many of our guests today talking about Nvidia being a long-term player
within their portfolio.
Angelo, Bob, Ivana, thank you all so much for helping us break this down.
Appreciate your time.
