Bloomberg Audio Studios, podcasts, radio news fourth.
Quarter revenue has beat estimates. Fourth quarter data center revenue coming in at thirty five point six billion dollars thirty four point oh nine billion dollars. Fourth quarter revenue that top line thirty nine point three billion dollars, exceeding estimates of thirty eight point twenty five billion dollars. Now coming
to that revenue estimate for the first quarter. In Video C's first quarter revenue plus or minus two percent, uh, the estimate was for forty two point twenty six billion dollars. Fourth quarter data center revenue coming in at thirty five point six percent, the company saying that it ramped up massive scale production of Blackwell AI. In the after hour shares hire right now by about three point four percent.
Yeah, actually in Video saying demand for Blackwell is amazing, So that's what we talked about in terms of that transition. But again the key metrics, as Tim said, stock is up almost four percent here in the aftermarket. The first quarter revenue, so looking ahead outlook forty three billion dollars plus or minus two percent. The estimate was for forty two point three billion, and again fourth quarter data center so looking at the quarter that they just wrapped up
data center revenue. That's the bulk of their business, that's those hyperscalers and more. Thirty five point six billion. That is a beat. Thirty four point zero nine billion was what the street was expecting first quarter, justin gross margin again looking forward seventy point five percent to seventy one point five percent. Street was looking for a little bit more. This is a measure of profitability, and the estimate on
the street was seventy two point one percent. Again, the stock now not up as much, In fact, it's down just a hair.
Yeah, shares climbed as much as four percent after this first quarter revenue forecast top estimates. We're getting some more commentary too, in Vidia saying that Blackwell is achieving billions of dollars in sales in the first quarter. The company also saying that it's successfully ramped up Blackwell massive production. In Vidia also coming out and saying, as Carol mentioned, demand for Blackwell is amazing. We're continuing to get more
numbers here too. Fourth quarter compute revenue coming in above estimates at thirty two point five six billion dollars. Fourth quarter networking revenue coming in just shy of estimates a three point oh two billion dollars.
Ye, now the stock is down about one percent here, it's bouncing around a lot, not surprising a lot of investors in this name. So you're going to see a lot of movement as we get tick by tick, headline by headline. But again, the key ones right now, it has to do with the fourth quarter data center revenue. Excuse me, thirty five point six billion. Again, that's above what the Street was expecting. Street was looking for thirty
four point zero nine billion. And again key is the revenue outlook forty three point zero billion plus or minus two percent. Street was at forty two point three billion.
And again we.
Got some context context, if you will, in terms of demand for Blackwell being amazing, successfully ramping up the Blackwell massive production, Blackwell achieving billions of dollars in sales in the first quarter and ramping up the massive scale production of Blackwell AI.
If the company else is saying it to lilivered eleven billion dollars of Blackwell architecture revenue in the fourth quarter.
Joining us right now, Red Brown is here in our New York studio, Bloomberg News Earnings reporter Caroline Hyde, co host of Bloomberg Technology, also here in studio. You know, Red, I want to bring you into this conversation. You keep a watch on earnings, just you know, your initial thoughts and what we're seeing in the trade because the stock is bouncing around a lot.
Yeah, I mean I do feel like we got to like kind of take a cyber relief here because they did check off all of the boxes kind of what Kun John answer the question marks Kun John has kind of touched on there. The outlook is solid, the fourth quarter was solid, and there seems to be real kind
of optimism around the progress with Blackwell. So yeah, I think overall this kind of was in line with what analysts were expecting, and the markets maybe been a little bit down going into this court into this report, but yeah, kind of a cyber relief I think for everybody.
The company did give an up eat forecast. It is a sign that the AI build out is strong. Carolin Heid, I want to bring you in, co host of Bloomberg Technology on Bloomberg TV Weekday at eleven am Wall Street time. I do want to ask about is there is the sigh of relief sort of coming too early? Did we get any answers about the impact of deep seek here.
Well, I think there's flies in the ointment.
I don't know how much you've talked about them, but I think the gross margin is clearly a sign that we're having perhaps a slight air pocket that many had talked about is where you flip from Hopper to Blackwell. And indeed, how much expense is spent on making sure that Blackwell is easy to use and get up and running as quickly as possible.
That could be an issue. And you have also got.
This potential guide of slightly weaker on the revenue for the fiscal first quarter plus or minus forty three billion. If they come into the low end, that does indeed just slightly lose out. But I think we are trying to hear ultimately whether that insane demand is now an amazing amount of demand and does that basically put off any concerns that we had about whether or not more
can be done with less in the future. And I think we'll hear on the call, particularly from Jensen, as to ultimately whether inferenced demand is really there for his very expensive chick tyrol.
It is like reading a statement from the Federal Reserve insane versus amazing.
This is what I was going to say ask you, is that you know you follow this company so closely, and it's how do you read the tea leaves because he does tend to be very enthusiastic and maybe changes the word and maybe there's a nuance to that change, But how do you kind of read through what he says, because we know he's going to be a beat.
I think he's going to be saying, look, trust us with this R and D, trust us on a slightly lower margin because we have got quote the fastest product ramp in our countries.
His company's history.
This is a company that suddenly become like Apple and giving us a new chip.
Every year that ever used to happen before.
There's an extraordinary pace which they're trying to innovate that costs money. And indeed, of course you have to be patient that sometimes there's going to be a slight friction when you come to actually putting these Blackwell supercomputers into the mix.
She shares, we're as high as four percent higher in the after hours, now down about four tenths of one percent. Yeah, not so much, like though investors are really trying to make sense of these numbers. Yeah, it's not going and it's not moving in a clear direction.
I think the clear direction is still always growth, isn't it? And the idea that sure and one hundred percent increase in revenue that got for their fiscal twenty twenty five, which is the one that's just gone, and the one prior to that, for two straight years we've had a doubling of revenue.
We're going to have to start to get used.
To a fifty percent increase in revenue and then a twenty three percent growth in revenue. The key question I think there is still going to be what about the cheap alternatives? What about Trainium too that Amazon has that is much cheaper, What about some of these startups that.
You're seeing, does Tridium too, does that compare with Hopper? Does it compare with black?
Well? About inference much more as well?
And so this is when they're trying to say, hey, you've got your beautiful large language model, come and now use it with us, and we will save you so much more money about thirty percent.
So after you've used Nvidia to train it, yes, we will process those requests for you exactly.
And I think that's where you're to start to get the edge.
And in fact, there was an interesting interview done with a rival with our CEO of course knows data centers and mobiles deeply, and his own design work of course helping a lot of those out there, including in Video, but he himself saying, look, this is where the opening is for competitors.
It's the inference stage. No one can really compete with the.
GPU power that in Video has when it comes to training these models. But the next layer, when we actually get it in and use these incredibly powerful things, that's where the edge can come.
I mean, I don't know. In terms of market expectations, not bad.
Yeah, signaling life kind of everyone was on the money, alas did their job well. Jensen's done his job well in terms of guiding. But had we got used to beaten raises. But look look back to when we were sit here together with Meta and Meta's numbers were a really late be in miss the forecast, and then Zuck gone on and just talked really optimistically and we all bought it.
Yeah that's a good point. Yeah, wait to see what happens in the cash session. Wait to hear from Jensen a little later. We can only take the moves as soon as the numbers come out with a grain of solid red brown.
I put deep seek in the search and nothing, nothing's coming up.
Let's just say it'll likely be asked about by at least one analyst tonight.
Red round.
I want to bring you in here on sort of the broader market reaction to this. There was this idea that given the segiment we've seen in the last five six days, uh in the broader market, the cell off that we've seen in the megacap tech companies this year and in the last few days, this would sort of change the tone, does it.
I mean, it doesn't really seem like the tone has changed at all. We're looking at the numbers is flat right, kind of across the board, with everybody not really reacting. You know, I think a lot of the narrative kind of has been said. You know, the spending is still there. I think if we read all the analysts reports heading into this, that was the one data point that everyone was hanging on to that there hasn't been anyone coming out and saying, you know, we're going to be adjusting
our AI spending. They've stuck to you know, those massive capex numbers that we saw throughout the quarter, so I think, yeah, maybe a little bit of a wait and see still unfortunately, you know, for the time being.
Well, and they did note I'm looking at our live blog noted margins are being squeezed by the cost of bringing new, more complicated products to market, and it sounds like what.
Well, Carolyn explained that means because you know Kunjohn also mentioned this idea of an air pocket or the transition from Harper to Blackwell. What exactly happens behind the scenes here and why is it so expensive?
Well, I think you've just ultimately got to make sure that it works hardware seamlessly with the software that you've got going, and ultimately that becomes more a powerful piece of equipment that is able to use your data and do what you want with it. I think when it came to Hopper, there was that was working beautifully for many, but everyone got deeply excited about the fact that there was the latest greatest innovation suddenly coming out and sort
of decided to go long on Blackwell instead. We actually saw a repositioning of that by many customers when they realized, oh, actually this is going to take work for everyone to ensure to pull this in, to get this into my data centers, and to ultimately make sure that.
It's doing what it wants to do with the most efficiency.
Heating problems for example, just some of the technicalities that they hadn't been able to stress test at this rapid rate of innovation.
All right, just to rehash, Nvidier giving a bullish revenue forecast for the current quarter, reassuring investors that spending on AI computing remains strung. Just reading some of the reporting out by our Oni and King, who follows this sector, sales will be about forty three billion in the fiscal first quarter, which runs through April. Analystaid An estimated about forty two point three billion on average, with some projections ranging as high as forty eight billion. So there were
some outliers and some very high expectations. Carolyn, how are you thinking about this story? Obviously we want to see what more commentary we get on the call the follow through. In terms of the trade, like I'm looking at NASDAQ one hundred mini futures, they're relatively little change, maybe up about one tenth of one percent. S and P five hundred mini futures, they are just a hair lower, pretty flat, So it's really very calm. It feels like you're in the aftermarket.
And maybe we just sort of braced ourselves because the volatility they experienced yesterday, and then say we got used to the fact that the market is trying to find its direction when it comes to what is the question of capacity, then we've got to go back to what happened in the beginning of the week, and we're all asking questions about TD Cowan's note over the weekend about Microsoft.
I'm sure that it will be questions put to Jensen about ultimately is there a capacity over stretch or underreach at the moment, and what about these so called scaling rules that he's trying to combat in every feat, whether or not it's about the scaling rules of how much better llms can become with the amount of data on compute you throw at it, but ultimately also the scaling rules of China or whether that becomes something you can still scale into in any way with an H twenty.
I'm glad he brought that up because one of the questions that kun John mentioned was getting more detail on views of more US restrictions, specifically with regard to China. Read I know you've been doing some searching within the press release. Are you seeing anything in there about China sales, about a commentary about China.
No mention about China. So then Kunjohn is correct that we that's definitely going to be a top of mind topic for analysts on the call. You know right now there's a lot of question marks looming around what they can do there. It gets they get around twelve percent of their sales over the past few years, most of that coming across from game being automotive end data centers.
So I think there will be questions about, you know, what sort of contingencies maybe they have in place if any sort of rediction restrictions do come down.
Well, and Caroline, I mean jensenmong has made his way to the way that's read. He said, conversations with President Trump about these issues and his concerns.
Yeah, certainly, and that had been pre premeditated. He'd been going to go and see President Trump. We know that he'd wanted to. When our own ed Ludlow sat down with him previously, he said, I'm ready willing and able to go and meet President Trump, and certainly they wanted
me to go. She because remembers the end of the administration, the Biden administration that's suddenly added all these curtailments to how much they can place chips in other geographies, not just China, because there's this worry that you sell into a different nation to Singapore, and actually that's suddenly how it's getting dispersed into China as well. So he's got
a lot to be debating. I do note in I mean, I hate to be a cynic when you're looking at these press releases, but he really only mentions US companies at the beginning of how much that they're working with the HWS is the Cisco's the verizons. I mean, it really is a focus in on us.
It's just this week things I think now increasing a partnership right with Cisco.
Exactly to get the networking better again, just get these things into people's hands more efficiently.
We've talked about hop Hopper, We've talked a lot about Blackwell. Is it too early for us to be talking about Reuben I don't.
I mean, what in the next innovation, This is the next year, this is the next I don't think Jadson is ever too early to talk about what he's got up next to sleep.
So that's that's the question. I mean, our investors looking for you know, this is this is the This is what happens with compute, right, You build something and then you build something better, and you build something better. That just happens over and over again, and it costs money. It costs money. So how important is you know the next generation what comes after blackwell apparently code named Rubin. How important is that to where we are in the timeline of Nvidia.
I think it's important to show that they are continuing to just innovate ahead of any of the competition. That the more it's more Lisa Sue who pays attention to Ruben and things up like, yeah, how am I managing to get any sort of in here in terms of owning the data center space when it comes to just the ability that Jensen has had to see around corners, when it comes to the use and application of his GPUs for gaming.
For crypto and then for generative AI.
But maybe Lisa Su starts to play the tack of okay, but it's me with cheaper price points, and I wonder whether the margin starts to creep into that narrative as well as maybe yes, it's a deeply expensive piece of kit with everyone offering cheaper alternatives.
What do you do?
How do you respond?
All right, folks, we're talking with Caroline Hyde, co host of Bloomberg Technology on Bloomberg TV at eleven am Wall Street Time. Let's bring into the conversation Mandy Seeing walking into our studio Bloomberg Intelligence, Senior Tech Industry Analysis. He walked in because we asked him to and we were hoping you would. Hey, and video shares have been bouncing around. Should any of us who are thinking about the AI industry? Is this a dose of optimism?
It is? But when it comes to Nvidia specific I would say, you know, the fact that networking revenue declines sequentially for the second quarter in a row again in Vidio is very high bar. And you know, given their bed and raises, everyone expects them to do, you know, well on every line. So networking going down, to me, that's a sticking point, along with gross margins going down and just the sequential increases that we have seen in the past quarters. It looks like it's tapering again. It's
double digit. But when it comes to the overall growth rate, it feels like it's decelerating. And the last data point I would offer is the cloud exposure, the hyperscale cloud exposure. It was forty five percent before now it's fifty percent Now. What it tells you is they kind of we are exposed. And so the fact that you know networking is going down, it's a sign that you know, training versus inferencing component that seems to be really in question here because networking
does well. If training is growing, if networking is going down, that's a sign that they may be facing some sort of challenges or starting to get more questions around the scale of training. And I think that's a big question, So go.
Deeper on that with training versus inference. We're talking a little bit about it with Caroline context Amazon and what Amazon has when it comes to in France. But what do we learn in video within Vidia today.
Look, if you're setting up a giant cluster like Xai has with you know, one hundred and two hundred thousand GPUs, you need a lot of networking there. But if you are setting up a data center with let's say a thousand, two thousand GPUs, you don't need as much of Nvidia as networking, and that's where you know, training versus inferencing really matters, because if you know, companies are leaning towards
setting up more data centers with fewer chips. I mean when I say, if you are still thousands of chips but not to the scale of one hundred thousand, then networking won't do as well. And that's the whole debate about training versus inferencing. So to my mind, you know, deep seek and all the recent developments from hyperscalers suggests that you may not need giant clusters and that is
I think going to affect Nvidia going forward. So when I look at their guide, when I look at the fact that you know, networking declined sequentially, again, it's a great print, but to me, the path seems to be that it will slow down on the networking side, and then you have to ask yourself where will the upside come from. Maybe on the call they expand on starlink and you know, the sovereigns that is a big pocket.
If that pans out in terms of training, like setting up big clusters that Sam Altman has talked about, you know, with projects started, then you will start to see networking bounce but for me, there is a clear connection between
networking and size of cluster and also software. There was nothing in the print around software, even though it's you know, two billion dollars in revenue when you look at a data center, you know, aligne, which with over one hundred and fifteen billion dollars two billion is very great small, but they didn't mention anything around software. So there's still a chip company that depends on you know, companies buying their their chips in bulk and also leering and networking.
So to me that the networking part is critical here.
You know, one of our on our live blog are Carmin Rhinikey saying, you know, bringing kind of salesforce into the mix, noting that shares of Salesforce are down more than five percent in post market trading after it reported we growth outlooks, stoking worry about its new AI product. Like you know, you do think about kind of the whole infrastructure, the whole community. Caroline, come on in the conversation following off mandypin if you've got a question.
Well, I do think that that's interesting about agent KI and this is where salesforce and agent force is in the front and center and look, Jensen pays lip service to it in the statement saying AI is advancing at light speed as a GENTKI and physical AI set the stage for the next wave, but he's got to keep talking about scaling laws when it comes to training. That is going back to a reasoning AI. He's trying to find new ways to galvanize us all to be excited.
But if the agents aren't selling that well as far as we thought about.
Is reasoning really that good for Nvidia? In my mind, if the pivot has happened towards more reasoning, huh, that's not a positive for Nvidia because when you're doing reasoning, you are doing more inference compute. And I mean, based on Jensen's past comments, Nvidia has a forty percent share in inferencing. Sixty percent is training, right, So how can it be better if you're doing more training, you know,
using that reasoning approach. So those are like very small nuggets, but I think in the grand scheme of things they matter.
You mentioned agentic AGENTIC AI. You got to explain what that is for people, especially in the context of what you talked about with Panos Pane earlier and what Amazon released today, and.
I think this is how ultimately starts being more helpful that a not only will your chatbot basically go away and take longer thoughts about things, thoughts quote unquote, you know, trying to or personify these these pits of tar and it is real, it is real, and person fall in love with it. But she if it's ass then is going to go off and start actioning stuff for you. Not only are they going to tell you the best restaurant, they go book it for you. They're going to potentially
start booking flights, paying for them. Agents can take ultimately still relatively basic needs, but run with them and complete the task. And that's ultimately where Salesforce and these enterprise software companies have started to really see. They're in the question for many which has is always being asked is
how they price for it? And Salesforce sort of stick the cell and sells on a limb and started saying, well, this is how much going to pay per per user or indeed per amount that you actually use.
But all of this is the return on AI that we keep questioning in so many ways.
I mean, you have to ask yourself, why is Nvidia's growth smart going down?
Right?
You know, obviously they release a new product that should command premium pricing, and look they guided to that. So they're saying in the second half you will see gross margins rebound, but they're going down for a reason. Either they're paying up more for foundry costs to DSMC or
they're just incanctivizing their customers to buy more. And you know, given black Belt did so well in the quarter, there could be some you know pricing aspect to it in terms of just driving demand and you know, shipping morels.
Is that because we're getting more competition as well in the future.
I mean CSPs, the four CSPs make fifty percent of their revenue hyperscalers, the hyperscalers.
So all you're talking about your alphabet, you're talking about your Microsoft microsycroso.
Yeah, and Amazon so and it was forty five percent last quarter, So that exporture grew, right, I mean, and they're shipping more blackwell. So when you connect it to clearly there is and these are customers that are buying chips in bulk, so they need some pricing, you know.
To Does that play to the margin that they're getting a break?
I would assume? So, right, Yeah, if you are a customer that is paying, you know, twenty billion dollars a year or two in Vidia, you want some pricing break compared to the other customers, not.
Just a free basket.
It was a question that I couldn't answer fully because I don't have the depth of knowledge that you do Man deep And ultimately, the issue of getting to Blackwell and the heating issues that we've seen and the ultimate rollout, has that cost them any margin? And does it cost them to have top going back and fixing things or is it on someone else to keep using ensuring it's working.
So there were concerns around Blackwell, and look, if you're a customer, you would want to wait until they resolve all these issues. But in Media moved at a pretty fast pace in terms of fixing those issues and actually getting eleven billion dollars of revenue in the quarter. So the fact that they were to do it so quickly
shows good execution. But they're shipping a lot of these chips to the hyperscalers, who again I feel, have to be incentivized to take those bulk shipments, and that could have played in terms of the lower margins.
And we're the concerns about overbuild answered where they put to rest.
I mean they look at all of their suppliers TSMC, Micron on the HBM side, all of them are still supply constraints, so I can't imagine, you know, there is any chance of an overbuild here. There's clearly more demand than supply and all the suppliers are continuing to expand capacity. So that is the least of my concern here. It's more the competition and the lower gross margins and the training versus inferencing is the key point that I hope he addresses on the call in relation to deep sea.
I'm just again going to people who maybe are not so in it, like you guys are the training versus inferencing, Like what do they need to kind of understand. I want to go back there because I do feel like that's the big factor here.
Building one million GPU clusters right now, the largest sized cluster is a two hundred K cluster that XCAI has built, and I think META is also a building, and.
That's for training or for enfortan training training.
You build a one million or one hundred thousand chip cluster, you're using it for training, You're not doing inferencing because inferencing patterns are very volatile, and inferencing is more tied to usage. You know, if I am using a chat gipt during the day, the traffic will be more, at night it will be less. So I'm not going to keep the chips under utilized. I want one hundred thousand
chips to be utilized at maximum capacity. That's why training is what is used for, you know, the large GPU clusters.
But to that point, even to the inference point we learn, I think today that deep seek is sort of doing surge pricing when it comes to when you're using the app because they want people to use it not during peak times because that is expensive.
Yeah, I mean look, and that goes to the under supply aspect, right, so the market is still under some and that's where wherever you can find capacity, you use it.
We know, you guys both have to go. Caroline, So you're sitting down with Jensen, what do you want to know?
I want to know about For me, I want to know about China, and I also want to know about this capacity issue that we keep talking about and how ultimately at Easy is with the amount that demand he still go going all.
Right, same thing for you, man, Deep, you're sitting down with Jensen.
Hey, right, training, Just tell me your training share right now? Is there still sixty forty? Has it changed? And what's the outlook?
All right?
Guys, you rock.
Thank you so much. Man Deep seeing Bloomberg Intelligence senior tech industry analyst here in studio along with Caroline Hide Coast of Bloomberg Technology and Bloomberg TV. Check out Man Deep's research. It'll be on the Bloomberg tomorrow. Caroline will be all over it on Bloomberg Technology. That comes your way at eleven am Wall Street time on Bloomberg TV.
All right, well, you heard Caroline mention China just now, and that's kind of what we want to talk about with Michael Shepherd. He's Bloomberg News senior editor for Technology and Strategic Industries. He joins us from our wash Rington, DC bureau. As Red Brown pointed out to us, Michael just a few moments ago, no mention of China in the press release. We are waiting for Jensen to take questions from analysts about that. What are the concerns around potential.
I don't want to use the term sanctions, but restrictions on what in Nvidia can sell to China.
Well, that's a better way of putting it, Tim, because really what in Vidia has confronted it's not some it sanctions or penalties, but limits on the types and quantities of chips that can sell to Chinese entities. And of course this is in response to the challenge that the world's second largest economy poses to the US in advanced technology, but especially in this emerging area of artificial intelligence. And Donald Trump upon taking office, he made it a national
priority to preserve dominance and artificial intelligence. But that was also a theme during the Biden administration, and some of these restrictions that have really been deviled in Vidia at times date to the Biden era and President Joe Biden, a week before leaving office and handing over the reigns of power to Donald Trump, imposed a three tiered system of limits on country's access to sensitive chips for AI
of the likes of which that in Nvidia produces. And so we are waiting to see what Jensen Wang will have to say about this, what sort of challenge as opposed to the business, and what sorts of steps has he taken then maybe try to talk Donald Trump out of it.
Has he been so has he been to the White House yet.
And he was one of the last of the big tech executives to pay a visit to the White House and see Donald Trump. There was no visit by Jensen Wang to mar Lago during the transition, and it wasn't until January thirty first, that week when deep seek really shook the markets, and especially in Vidia. It was a Friday that he went in to go meet with the President and talk generally about artificial intelligence, of course, but you had to imagine that export controls came up. Nvidia
certainly has railed against them. We saw statements coming from their spokespeople, as you know, we heard the Biden folks talk more about them. And then as the Trump team came in and saw a deep Seek's breakthrough and thought, hey, we have to do something ourselves. We may need to further tighten these restrictions. Well they weighed in again. But Jensen Wong has been quiet about it. So this will be a moment for him to say something.
You know, I think it was Caroline, I can't remember exactly who it was on our on our Eric Carrol who mentioned that in the press release, the names that are mentioned cloud service providers aws Core Weave, Google Cloud Platform, Microsoft Aser, Oral Cloud, Oracle Cloud Infrastructure, AWS, Cisco companies, all US companies, And I don't know if it's if it's a political statement or if it's just a statement of the fact that these most prominent companies are based
in the US. Is there a read into this, Michael, that this is a company that's making sure it's touting its American connections.
So many companies, Tim McCarroll, are trying to get on the right side of Trump in that sense, and in this case, the facts may be conveniently aligned to whatever Trump and the new administration may want to hear about American companies buying from an American provider, namely in Vidio. But of course, so many of Nvidia's chips are made in Taiwan by TSMC, so that has to be another area of concern for Jensen Wong as he hears Donald
Trump talk about potential tariffs on semiconductor imports. He hasn't imposed him aet he hasn't really announced him yet, but it is something that we are watching for in the coming weeks, and of course, as it will be delivering this address to Congress next Tuesday, evening, and that could be a moment where we hear him talk in somewhat more detail about this. Remember, the semiconductor supply chain is
so interconnected globally. You have foundries in Taiwan making chips that are designed here in the US, and and so all the production and design are happening in a lot of different places. So when you see tariffs and export controls imposed, you know, the trade flows and the impacts really ricochet through the supply chain and all the way through.
You know, when you get to end use products like computers and like cars, you know where we're going beyond the data centers that we think of with Nvidio, But speaking more broadly on the semiconductor sector, the ripple effect through the supply chain will be really strong.
Michael.
And you know, certainly something you guys and you are, you know, on top of on a regular basis in terms of things coming out of the White Houspital. Do you wonder about when it comes to ai and what President Trump wants to see. Obviously he wants to protect the US industry, the US chip industry, but so many of these big tech companies here in the United States are global players. It's a fine dance, right in terms
of their top line, their revenue, their growth. Yes, the US market obviously important, but global markets are important as well.
That's a great point, Carol, because when you think about one of the best known US chip companies, it also happens to be one that is struggling existentially right now, and that is Intel. And it has sales overseas, but including in China. So any export restrictions, any move that could affect its ability to do business in China could be bad for Intel and in Vidia as well. You know the You know, it's not like Nvidia has no business in China, but the most sophisticated chips are or
off limits. Now the administration is wighing weather to lower the threshold on those limits, making even less advanced ships made by Nvidia and other makers too off limits for Chinese buyers, barring a license of permission from the US Commerce Department. So they are thinking about this, and we're looking at a series of restrictions too, and these include getting Japan and the Netherlands to get companies like Tokyo
Electron and ASML. These are the businesses that make the semiconductor manufacturing equipment, these huge photo lithography machines that are really finicky and expensive. What they want is they want Japan and the Netherlands to limit maintenance of these machines that have already been sold to China. You know, they're looking at every which way to contain China in this area.
Michael, hang on for a second. I just want to recap because we are seeing shares of Nvidia up about two point two percent now in the postmarket trade, after market trade. They've been bouncing around a lot gains and losses, but it does look like they're kind of holding at
this level again, a two point two percent gain. Keep in mind, I'm looking at our Ian King his reporting and video giving solid quarterly results and a bullish revenue forecast for the current period, even if the numbers didn't reach the blowout level that some investors were banking on. And of course we're waiting for some headlines tim here as the call will get underway soon with the C suite and Jensen Wong of Nvidia so big.
I think the big question that I have is how much Insight will get And we don't know this yet, Michael, but how much Insight will get on the call into invidious relationship with China and to what extent restrictions are set to affect the company.
And that is a key question for investors going into this because when Deep Seek emerged onto the scene, and remember Deep Seek was a trigger for the Trump administration to think about tougher restrictions on Nvidia's chips, and it
was also an existential moment for in Vidia itself. You know, Deepseeak was able to produce its chatbot on the cheap, undermining this whole idea that hey, we need Manhattan Project level CAPEX to be able to produce these breakthroughs in AI, and of course to do that you need tons and tons of Nvidia accelerators. Deep Seek raise questions about, well,
maybe we don't need quite so many work quite so much. Nonetheless, we've seen the hyperscalers all, you know, reinforce their commitment to heavy CAPEX tens and tens of billions of dollars in plans for data centers and chip purchases.
All right, So appreciate being able to get this perspective and roll it in. It's such a big part of the Nvidia story. Michael Sheppard, thank you, as always, Senior editor for Technology and Strategic Industries out there in our Bloomberg News Bureau in DC,