00:00:00
Speaker 1: Bloomberg Audio Studios, podcasts, radio news.
00:00:09
Speaker 2: This is a breaking news update from Bloomberg instant reaction and analysis from our three thousand journalists and analysts around the world. We've got a great group of people joining us to talk all things in video, including our own ed Ludlow. He's the co host of Bloomberg Tech. He joins us from our San Francisco bureau, kun John Sabanni's Bloomberg Intelligence. Senior semiconductor analyst, Paul Meeks is also with us, head of Technology research at Freedom Capital Markets, and Jay Goldberg of Seaport Research Partners is joining us as well. Once again, Jay is the only analyst tracked by Bloomberg who has a cell rating on in video stock. He's got a price target of one hundred and forty dollars. Paul, your reaction please with these numbers.
00:00:58
Speaker 3: It seems to me that everything is in order.
00:01:01
Speaker 4: Like everybody else in the preview commentary, the most important metric that they could release ahead of a call was this revenue guide, and it looks fine. I'm a little bit, i don't want to say dubious, but a little bit confused by the change in the reporting structure.
00:01:18
Speaker 2: Why would a company do that, a company like a video that has such a following right now, so many analysts covering this. Some people considered a macro event when it reports earnings, why change reporting structure right now?
00:01:33
Speaker 4: You know, one of the things that they are doing is they're trying to build a walled garden, just like Apple did in a previous technology transition, by going beyond the GPUs and interconnecting in the data center all the different pieces. So this may not be some sort of dodge, you know, because the company typically is first class, very transparent, but it might be just refor that they are moving beyond just selling GPUs and they want to organize their business to prove it.
00:02:08
Speaker 5: Kun John, let's bring you back in here, Bloomer Intelligence senior semiconductor analyst, and just get your initial thoughts on these first few numbers.
00:02:19
Speaker 6: Yeah, look, everything looks closier like your earlier guest was saying. You know, they have made us habituated to expect them to come out and expect and to beat by such a range mid single digits. And that's a problem. When you do that, it has become sort of a baseline, no longer something that stands out if they didn't beat with this magnitude, that would be something worse. I do like the buyback announcement and to address that resegmentation.
00:02:44
Speaker 1: I think we like that move.
00:02:46
Speaker 6: I mean, look, our model has been trying to guestimate this breakdown that they're giving in the data center business by the type of end customers for a while now. It's always painful to really get that out of them and figure it out. So we like that move that now they are coming up and disclosing that breakdown.
00:03:04
Speaker 3: That also allows.
00:03:05
Speaker 6: People to really see that it's not just the four top cloud customers that they're buying and running the business for in media right, there's also growth coming from other vectors that people often miss out because it was not publicly.
00:03:18
Speaker 1: Showcased and ed.
00:03:20
Speaker 5: We are just getting a headline crossing that the company is saying that no shipments of data center Harper products to China. How does that live up to the expectations of what you and Jensen Wang had discussed.
00:03:35
Speaker 7: How you did that so quick? I literally just posted that in the blog before before we read it out. This is recently, that's the base case assumption. I mean, what collect cress, the CFO is pointing out literally in her CFO commentary as opposed to figuratively, is that this is the fiscal first quarter of financial year twenty seven. You know, the base case assumption was no product data center wise to China, so it meets that expectator reminding me in that note that in the first fiscal quarter of the fiscal year twenty six they had almost four point six billion dollars of sales in that same period one year ago. So that's all we're learning. Really. There were some other stuff that I scanned through on the revenue highlights of the CFO commentary. I thought we'd learn more, but right now, that's it.
00:04:20
Speaker 2: What's the question you'd ask about that on the call if you were given the chance.
00:04:23
Speaker 7: Ad why did you tell everyone in March that you were ramping up your supply chain and that you had both orders from Chinese companies along with the US licenses to issue those orders. Why'd you say that, you know, if the state of play is clearly based on the President's visit, that China is not going to permit its tech companies right now to place those orders. And why did ed Ludlow not ask Jensen that more explicitly on Monday. That's probably what I would reflect on.
00:04:52
Speaker 2: Hey, you know you can even Monday morning quarterback your own work and pretty easy to do that. It was a great interview with Michael dell as well too from Las Vegas, want to bring back, Paul Meeks, Heat of Tech research at Freedom Capital Markets, and also Kun John Sabani Bloomberg Intelligence senior semiconductor analysts. They're joining our roundtable. Coon John, you mentioned that you like this additional eighty billion dollars in share buyback. Why why is that a good use of cash? Like convince people that that's a good use of cash in your view?
00:05:19
Speaker 6: I mean, given the stage VR in like I said, like you know, every quarter now, in terms of near term numbers, it's basically table stakes that to beat by mid single digit. They've already given a two year pipeline of a trillion dollars, So I don't know what else they could do just in terms of the print to really get excitement. So buy back usually is a good strategy. And remember they're generating enormous amount of cash that right now, despite all the big investments they're making, massive multi billion dollar investments they're making, they don't have other way what to do with that cash. So a buyback always is sort of preferred go to for US analysts that does the reward shareholders from time to time.
00:06:00
Speaker 5: Well, I do want to get your thoughts on just the market's initial reaction. Actually our Bloomberg coverage the headline in Vidia gives disappointing forecast as chip competition mounts. So we do know that number ninety one billion, but disappointing.
00:06:17
Speaker 8: What do you make of that word?
00:06:20
Speaker 3: I think it's bs. Yeah, I think it's just fine.
00:06:25
Speaker 4: And we'll get some more commentary as we get into the call starting at five o'clock Eastern. But now the stock is trading after ours flatish, and you know, the stock it had a pretty big run up in recent weeks after it essentially was a poor performer from July of twenty five until just a few weeks or months ago. And so I don't think that there's any problem in all the guidance. I think if they do announce Chinese business, take whatever revenues you're expecting for that fiscal year and add fifteen percent to it, so it's only going to be icing on the cake.
00:07:05
Speaker 3: But I, yeah, I feel pretty good about it. But you know Jensen will do what Jensen does right.
00:07:10
Speaker 4: He's the Ai Jesus and he's an ambabudlist and make us all feel better about the long term. I'm not necessarily worried about what they printed or even this guidance, But this guidance is not disappointing. I think it's fine because you know what, It'll be ninety one billion dollars and then they're going to upside it like they always do, so this will be much better at the end of the day than the whisper.
00:07:32
Speaker 2: We're speaking to Paul Meek's head of technology research at Freedom Capital Markets, also with us Kun John Sabani, Bloomberg Intelligence Senior semiconductor analyst and own Ed Ludlow, a Bloomberg Tech co host. Holding down the Fort in the San Francisco studio, the company gave investors a sales forecast, saying that revenue in the three months ending in July will be ninety one billion dollars, the company said in a statement, though analyst estimated eighty seven billion. On average, projection search as high as ninety six billion dollars, according to data compiled by Bloomberg. I want to bring back Kun John Sabani, Bloomberg Intelligence, senior semiconductor analysts. I know, Coon, John, You've got to run in just a couple of minutes. What's your note going to be headlined tomorrow?
00:08:13
Speaker 6: From the press release? I would say, everything looks fine, no need to panic, no need to worry, and the stock seems to be turning around, by the way, as we are going through this roundtable, all.
00:08:22
Speaker 2: Right, Kun, John Sabani's got to run. He's Bloomberg Intelligence, the senior semiconductor analyst. Our roundtable, though continues. We got Paul Meeks, had of Technology research at Freedom Capital Markets and Ed Ludlow, Bloomberg Tech co host. Hey, Ed, I want to go back to you as you've had a little bit of time to digest this and actually read some of the coverage too from a big team here at Bloomberg News that's reporting on this. What continues to stick out to you from these results?
00:08:46
Speaker 7: Yeah, you put the numbers to one side, right, So ninety one billion dollars plus or minus two percent in the physical second quarter. You know, it puts that outlook or that guide pretty firmly in the bullish part of the consensus or forecast on sell side and buyside expectations were very high.
00:09:03
Speaker 8: So what's the extra bit here?
00:09:04
Speaker 7: What is beyond in Vidia selling GPUs at the level that we would expect, And if you read collect Cres the cfo's commentary, it's that extra content in the data center, Infinniban Networking Spectrum, x Ethernet, envy link Systems, which is what combines a lot of the servers in aggregate to form that supercluster or supercomputer that seems to maybe account for some of the beat in the court had gone, and the confidence in the outlook to come, and you know, we are bouncing around now. We're modestly high two tents of one percent. We were lower. I think your last guest, Paul himself said right like, you're not going to get everything from the earnings release. Let's wait and see what Jensen Wang has asked and what he says.
00:09:45
Speaker 8: Paul, I do want to go back to.
00:09:49
Speaker 5: Just kind of on some of your thoughts on the company's data center expansion, because in Vidio had announced a few weeks ago investing as much as two point one billion dollars in the data center developer iron or I'm not sure is it iron or Ern? You have a hold rating on that company. What are you kind of looking for for Nvidia to talk about with that partnership.
00:10:10
Speaker 4: I don't know if they'll specify it because in Vidia, because it generates so much free cash flow. As you may have noticed the last couple of weeks and months, they seem to be sprinkling the infield with two billion dollars for everybody like Oprah if they do highlight IRN know IRN is a neo cloud. And what they did with Iron it was pretty unique because first of all, they bought some capacity for Nvidia's own use, not to resell to customers. And then they also don't necessarily have a warrant, but they have a right to purchase more Iron stock and actually a significant amount, but it's tied to Nvidia in IRN delivering GPUs together, and I think they get the full payout once they've supplied Iron and its customers six hundred thousand of their GPUs.
00:11:00
Speaker 3: So it is a unique deal. It is a little bit different.
00:11:03
Speaker 4: But in Video, with all of its cash, it's been doing two things. It has been acting like a venture capitalist. The plant seats throughout the entire ecosystem. Remember my point that they're going beyond GPUs, and because they don't really have anything else to spend down, they're buying back.
00:11:18
Speaker 5: Some stock ed Lodo. You interviewed the Iron CEO on Monday. What was your take on the company's plans for this kind of partnership.
00:11:29
Speaker 8: Yeah, so, you know, co CEO, Daniel Roberts.
00:11:32
Speaker 7: You know, Iron is a neocloud that is largely dependent on video architecture. And what is so interesting is we're talking about the industry being in a position where demand is fast outpacing the industry's ability to supply. But you know, as Iron would tell it, the bottleneck is not in video being able to get the server designs or the GPUs into place. It is everything else. It is the utilities trying to get the instruction place to switch that data center online. Is the concrete, the manual labor, the tin can that needs to go around it. So it's very interesting to hear that experience through through fair experience of what it's like to try and build. And Neocloud's just a data center for away AI workloads, simple, right, but in video doesn't appear to be the hold up here.
00:12:20
Speaker 2: Let's bring back Jay Goldberg Seaport Research Partners analysts there, he again is the only analyst track by Bloomberg with a CEL rating on in video. Jay, I imagine that this report is not enough and I don't want to get in trouble with compliance or anything, but not enough for you to change your rating on in video.
00:12:40
Speaker 9: Nothing I've seen here sort of meaningfully alters the trajectory of how I think of in video's well you prospects.
00:12:48
Speaker 2: What is your bear what is your bear case on the company, and how do you see it? How do you see the stock falling so much in just the next twelve months.
00:12:56
Speaker 1: To just be clear, I've never told people to short in video.
00:12:58
Speaker 9: I'm always positioned I might take as an underperformed go back to when you interviewed me in May of last year. This is that was what I'm saying, And I think that's that's played out right. It is it's just underperformed. It's it's it's a good company, but it's hard for them to live up to, you know, five trillion dollar market cap expectations, right, and and we've seen that this year, like they've underperformed the socks and the SMH ETFs underperforming, they've they've gone up whatever, twenty thirty percent this here, I think, and lots of other names in the sector have gone up a lot more, and it's just a function of how big they are.
00:13:34
Speaker 1: And I think there's if you look at the results.
00:13:36
Speaker 9: I mean, one of the things that stood outs to me was they're increasing the dividend, which is a great signal, but it's a massive increase, and right, it's sort of signaling that they're running out of ways to grow faster, Like couldn't they've used that those funds in a different way.
00:13:52
Speaker 2: I just want to point out you've accelerating on the stock. One hundred and forty dollars is your price target, and that is a period of twelve months. Yes, okay, so you're not telling people to short the stock, but you don't. But you believe that the valuation should be one hundred and forty dollars per share, not two hundred and twenty dollars a share.
00:14:10
Speaker 9: Yeah, I'll obviously revisit that today, but certainly I think it's I think it's hard. It's hard to get excited about the stock when they are better opportunities out there.
00:14:19
Speaker 2: Paul, come on in. I imagine you disagree with Jay. There is it. Is it easy for you to get excited about in video still.
00:14:27
Speaker 4: Well, Jay had some very good points, and I obviously admire him and his work, but I would tell you that this is not a situation like Cisco was right before the Internet bubble popped, where it was trading in a couple hundred times earnings. If you believe, and I don't formally cover in Vidia, I cover the Neo clouds, But if you believe the consensus estimate for next FISC year now, this would be the fiscal year for Invidia that ends January thirty, first of twenty eight, which is really calendar twenty seven. Right now, we are trading at about half of the growth rate expected in that year. Well, Jenny had some very good points, and I obviously admire him and his work, but I would tell you that this is not a situation like Cisco was right before the Internet bubble popped, where it was trading at a couple hundred times earnings. If you believe, and I don't formally cover in Nvidia, I cover the Neo clouds, But if you believe the consensus estimate for next FISK year, now, this would be the fiscal year for Nvidia that ends January thirty first of twenty eight, which is really calendar twenty seven. Right now, we are trading at about half of the growth rate expected in that year, and we're trading at an S and P five hundred multiple. Actually I calculate about one click below that. And here's a company that really hasn't seen much multiple expanse over the years. It's all been in delivering the mail. And so my point is, I'm not necessarily disagreeing with the trend longer term, and it is hard to grow something this large, but we're treading it less than a market multiple. It's not actually expensive at all. If it was expensive, then I probably swing from bulla bear.
00:16:20
Speaker 5: But it's cheap, and I do want to bring you back in just going back to the China. The China story here, because we've seen this company's revenue, the revenue that they generate from China declining quarter after quarter. They're now continuing to exclude China revenue from their outlook for the second quarter. Does Nvidia need business from China at all?
00:16:47
Speaker 7: So there's the basics the numbers, which is the quarter did not include any Hoppa based systems to China. And the outlook for the current fiscal second quarter in financial year twenty seven, just as you said, does not reflect any China revenues or shipments to China. And what Jensen Huang in collect cress the CFO of said is that China from a starting pace of zero baseline zero is a fifty billion dollar opportunity. But if you if you again get out of the income statement, the story with China was a critically important end market for data center and as Jensen Wang would like to say, often fifty percent of the world's AI research or the researchers themselves are in China or from China. And so you know, if you buy Jensen Wang's mile high bigger picture view of what is going to happen with AI, the build out of AI training and transition to running of systems with inference to their mind, being a participant with China is critically important. And again you know in that interview Monday, and I'm sorry to go back to it, but you know the word he uses, he uses deliberate. The demand from Chinese tech companies is quote incredible. It's just they're not permitted by the Chinese government to do business within video.
00:18:03
Speaker 2: Yeah, certainly on the call a lot of questions and perhaps some more answers when it comes to the China business for Invidia. I want to bring back and remind everybody that was just Ad Ludlow. He's the tech Bloomberg Tech co host. You can catch him every day at Bloomberg TV at eleven am Wall Street time. Jay Goldberg Seaport Research Partners joins us too, along with Paul Meeks, head of Technology research at Freedom Capital Markets. Jay, I want to send a question over to you about the reporting structure and the reporting framework for in Video. I don't know if you had a chance to see this yet, Bart, this cut adds attention. It caught our attention when these numbers crossed at four twenty. Lin Dwan from our Bloomberg Technology team, she's a tech editor. She writes that the company's transitioning to this reporting framework that in the future, in Video will report for two separate market platforms, data center and edge computing. The company says the data center segment will include business with the hyperscalers and those building AI focused data centers. Edge computing will highlight the devices itself for PCs, game consoles, robotics, and the like. I know this is a little bit of a wonky sort of accounting question, but as an analyst like I know you want to see clearly defined areas when it comes to segment reporting. What's your what's your initial reaction to this new reporting structure.
00:19:20
Speaker 9: So at one level it makes sense. I have to think through it. But data center is really where people are focused on.
00:19:26
Speaker 2: It's kind of the number that only any It's the only number of people care about, right.
00:19:29
Speaker 9: It's the only number of people care about, with the exception of my my fourteen year old son, who only cares about the gaming GPUs.
00:19:35
Speaker 1: And it's going to be very sad, saddened by this news.
00:19:37
Speaker 2: Okay, they'll still get the gaming GPUs, they just won't be reported out separately.
00:19:42
Speaker 9: It's yeah, it's getting harder to get them, and they're not getting updated as fast as he likes.
00:19:46
Speaker 2: But well, that's interesting because that's the DNA of the company.
00:19:50
Speaker 1: That's the DNA of the company.
00:19:51
Speaker 8: Yeah, Ed, you're a gamer, come on in here.
00:19:54
Speaker 1: I'm a gamer.
00:19:55
Speaker 9: I'm a gamer, but your gamer two j I'm willing to forgive them on their piece of gaming GPU looks upgrades.
00:20:02
Speaker 2: And I know this is not the number that everybody's focused on. But since this is in videos d in videos DNA, and you're a gamer, what do you make of the gaming GPU issue? They're leaving it to someone else, Now what is it?
00:20:14
Speaker 8: I don't know that it's an issue.
00:20:16
Speaker 7: I mean, you know again, I'm just you know, in videos changing the reporting structure. I get that. So it's just not highlighting gaming. But gaming is not the growth engine. It's tiny for this company.
00:20:31
Speaker 8: It's it's tiny.
00:20:32
Speaker 7: But like again, maybe you just consider the GPU that goes into a gaming console or your you're home rigged gaming PC.
00:20:42
Speaker 8: It's still at the edge.
00:20:44
Speaker 7: What people say is at the edge you're running computation at a device local to you. And you know the story that jensmong has been telling on the road, right, I think CEES or GCC, whatever, whatever it is, whatever event, it's that the role of the GPU in the video games industry is now as much about the development of that video game, its ability to generate pixels, right, rather than animate That's the story for in video with gaming. The gaming industry is in Nvidia, GPUs and data centers still as much as it is about the GPU. And by the way, I got like this Frankenstein's Monster ASA. It's got an amazing and video GPU in it and an Intel processor and core. But like, so what man, It's all about the data center.
00:21:30
Speaker 2: It is all about the data centers. Hey, I'm speaking of the data centers. I want to bring Paul Meeks back in here. Ian King and Dina Bass, writing separately on the live blog, Ian talking about Hyperscalers still being the biggest providers of sales for that unit, but others a clouds industrial enterprise.
00:21:46
Speaker 1: Grew quicker quarter over quarter.
00:21:48
Speaker 2: It's a trend that investors will want to see accelerate to remove the concern that it's too dependent on customers who have their own components and an incentive to move away from Nvidia. Dina Bass weighing in and saying that at the same time time, the greater growth in non Hyperscaler customers does point to the idea that indeed, folks like Google and Amazon, while they remain huge in VideA buyers, are trying hard to expand the workloads they can put on their own chips and get more of their cloud customers to choose them over in Vidia GPU. You cover the neo clouds you understand what the hyperscalers want. What do you make of that sort of revenue diversity question? And the companies that have their own ability to do this pushing their customers into their own products.
00:22:31
Speaker 3: So I agree with the push and pull here.
00:22:35
Speaker 4: Obviously it's important that over time that enterprises go direct, not through hyperscalers, to broaden this market. I think it's a very positive trend.
00:22:48
Speaker 3: In Vidia, of course, will be a beneficiary.
00:22:51
Speaker 4: On the other hand, particularly this announcement the other day with Google. Yes, Google and Amazon in particular working very hard on their homegrown chips, and it looks to me like they don't want to just lower their own costs by using their own chips internally. But in Google's case, they're going to sell these chips on the merchant market, and so over time it could be a threat. Why because Google has very deep pockets, almost endless resources to compete when you have other companies that say they can compete over time with Nvidia, be careful because these guys have the best balance sheet there ever was. But Google probably can make a go of it, and I imagine Amazon will be on their heels
Speaker 1: Bloomberg Audio Studios, podcasts, radio news.
00:00:09
Speaker 2: This is a breaking news update from Bloomberg instant reaction and analysis from our three thousand journalists and analysts around the world. We've got a great group of people joining us to talk all things in video, including our own ed Ludlow. He's the co host of Bloomberg Tech. He joins us from our San Francisco bureau, kun John Sabanni's Bloomberg Intelligence. Senior semiconductor analyst, Paul Meeks is also with us, head of Technology research at Freedom Capital Markets, and Jay Goldberg of Seaport Research Partners is joining us as well. Once again, Jay is the only analyst tracked by Bloomberg who has a cell rating on in video stock. He's got a price target of one hundred and forty dollars. Paul, your reaction please with these numbers.
00:00:58
Speaker 3: It seems to me that everything is in order.
00:01:01
Speaker 4: Like everybody else in the preview commentary, the most important metric that they could release ahead of a call was this revenue guide, and it looks fine. I'm a little bit, i don't want to say dubious, but a little bit confused by the change in the reporting structure.
00:01:18
Speaker 2: Why would a company do that, a company like a video that has such a following right now, so many analysts covering this. Some people considered a macro event when it reports earnings, why change reporting structure right now?
00:01:33
Speaker 4: You know, one of the things that they are doing is they're trying to build a walled garden, just like Apple did in a previous technology transition, by going beyond the GPUs and interconnecting in the data center all the different pieces. So this may not be some sort of dodge, you know, because the company typically is first class, very transparent, but it might be just refor that they are moving beyond just selling GPUs and they want to organize their business to prove it.
00:02:08
Speaker 5: Kun John, let's bring you back in here, Bloomer Intelligence senior semiconductor analyst, and just get your initial thoughts on these first few numbers.
00:02:19
Speaker 6: Yeah, look, everything looks closier like your earlier guest was saying. You know, they have made us habituated to expect them to come out and expect and to beat by such a range mid single digits. And that's a problem. When you do that, it has become sort of a baseline, no longer something that stands out if they didn't beat with this magnitude, that would be something worse. I do like the buyback announcement and to address that resegmentation.
00:02:44
Speaker 1: I think we like that move.
00:02:46
Speaker 6: I mean, look, our model has been trying to guestimate this breakdown that they're giving in the data center business by the type of end customers for a while now. It's always painful to really get that out of them and figure it out. So we like that move that now they are coming up and disclosing that breakdown.
00:03:04
Speaker 3: That also allows.
00:03:05
Speaker 6: People to really see that it's not just the four top cloud customers that they're buying and running the business for in media right, there's also growth coming from other vectors that people often miss out because it was not publicly.
00:03:18
Speaker 1: Showcased and ed.
00:03:20
Speaker 5: We are just getting a headline crossing that the company is saying that no shipments of data center Harper products to China. How does that live up to the expectations of what you and Jensen Wang had discussed.
00:03:35
Speaker 7: How you did that so quick? I literally just posted that in the blog before before we read it out. This is recently, that's the base case assumption. I mean, what collect cress, the CFO is pointing out literally in her CFO commentary as opposed to figuratively, is that this is the fiscal first quarter of financial year twenty seven. You know, the base case assumption was no product data center wise to China, so it meets that expectator reminding me in that note that in the first fiscal quarter of the fiscal year twenty six they had almost four point six billion dollars of sales in that same period one year ago. So that's all we're learning. Really. There were some other stuff that I scanned through on the revenue highlights of the CFO commentary. I thought we'd learn more, but right now, that's it.
00:04:20
Speaker 2: What's the question you'd ask about that on the call if you were given the chance.
00:04:23
Speaker 7: Ad why did you tell everyone in March that you were ramping up your supply chain and that you had both orders from Chinese companies along with the US licenses to issue those orders. Why'd you say that, you know, if the state of play is clearly based on the President's visit, that China is not going to permit its tech companies right now to place those orders. And why did ed Ludlow not ask Jensen that more explicitly on Monday. That's probably what I would reflect on.
00:04:52
Speaker 2: Hey, you know you can even Monday morning quarterback your own work and pretty easy to do that. It was a great interview with Michael dell as well too from Las Vegas, want to bring back, Paul Meeks, Heat of Tech research at Freedom Capital Markets, and also Kun John Sabani Bloomberg Intelligence senior semiconductor analysts. They're joining our roundtable. Coon John, you mentioned that you like this additional eighty billion dollars in share buyback. Why why is that a good use of cash? Like convince people that that's a good use of cash in your view?
00:05:19
Speaker 6: I mean, given the stage VR in like I said, like you know, every quarter now, in terms of near term numbers, it's basically table stakes that to beat by mid single digit. They've already given a two year pipeline of a trillion dollars, So I don't know what else they could do just in terms of the print to really get excitement. So buy back usually is a good strategy. And remember they're generating enormous amount of cash that right now, despite all the big investments they're making, massive multi billion dollar investments they're making, they don't have other way what to do with that cash. So a buyback always is sort of preferred go to for US analysts that does the reward shareholders from time to time.
00:06:00
Speaker 5: Well, I do want to get your thoughts on just the market's initial reaction. Actually our Bloomberg coverage the headline in Vidia gives disappointing forecast as chip competition mounts. So we do know that number ninety one billion, but disappointing.
00:06:17
Speaker 8: What do you make of that word?
00:06:20
Speaker 3: I think it's bs. Yeah, I think it's just fine.
00:06:25
Speaker 4: And we'll get some more commentary as we get into the call starting at five o'clock Eastern. But now the stock is trading after ours flatish, and you know, the stock it had a pretty big run up in recent weeks after it essentially was a poor performer from July of twenty five until just a few weeks or months ago. And so I don't think that there's any problem in all the guidance. I think if they do announce Chinese business, take whatever revenues you're expecting for that fiscal year and add fifteen percent to it, so it's only going to be icing on the cake.
00:07:05
Speaker 3: But I, yeah, I feel pretty good about it. But you know Jensen will do what Jensen does right.
00:07:10
Speaker 4: He's the Ai Jesus and he's an ambabudlist and make us all feel better about the long term. I'm not necessarily worried about what they printed or even this guidance, But this guidance is not disappointing. I think it's fine because you know what, It'll be ninety one billion dollars and then they're going to upside it like they always do, so this will be much better at the end of the day than the whisper.
00:07:32
Speaker 2: We're speaking to Paul Meek's head of technology research at Freedom Capital Markets, also with us Kun John Sabani, Bloomberg Intelligence Senior semiconductor analyst and own Ed Ludlow, a Bloomberg Tech co host. Holding down the Fort in the San Francisco studio, the company gave investors a sales forecast, saying that revenue in the three months ending in July will be ninety one billion dollars, the company said in a statement, though analyst estimated eighty seven billion. On average, projection search as high as ninety six billion dollars, according to data compiled by Bloomberg. I want to bring back Kun John Sabani, Bloomberg Intelligence, senior semiconductor analysts. I know, Coon, John, You've got to run in just a couple of minutes. What's your note going to be headlined tomorrow?
00:08:13
Speaker 6: From the press release? I would say, everything looks fine, no need to panic, no need to worry, and the stock seems to be turning around, by the way, as we are going through this roundtable, all.
00:08:22
Speaker 2: Right, Kun, John Sabani's got to run. He's Bloomberg Intelligence, the senior semiconductor analyst. Our roundtable, though continues. We got Paul Meeks, had of Technology research at Freedom Capital Markets and Ed Ludlow, Bloomberg Tech co host. Hey, Ed, I want to go back to you as you've had a little bit of time to digest this and actually read some of the coverage too from a big team here at Bloomberg News that's reporting on this. What continues to stick out to you from these results?
00:08:46
Speaker 7: Yeah, you put the numbers to one side, right, So ninety one billion dollars plus or minus two percent in the physical second quarter. You know, it puts that outlook or that guide pretty firmly in the bullish part of the consensus or forecast on sell side and buyside expectations were very high.
00:09:03
Speaker 8: So what's the extra bit here?
00:09:04
Speaker 7: What is beyond in Vidia selling GPUs at the level that we would expect, And if you read collect Cres the cfo's commentary, it's that extra content in the data center, Infinniban Networking Spectrum, x Ethernet, envy link Systems, which is what combines a lot of the servers in aggregate to form that supercluster or supercomputer that seems to maybe account for some of the beat in the court had gone, and the confidence in the outlook to come, and you know, we are bouncing around now. We're modestly high two tents of one percent. We were lower. I think your last guest, Paul himself said right like, you're not going to get everything from the earnings release. Let's wait and see what Jensen Wang has asked and what he says.
00:09:45
Speaker 8: Paul, I do want to go back to.
00:09:49
Speaker 5: Just kind of on some of your thoughts on the company's data center expansion, because in Vidio had announced a few weeks ago investing as much as two point one billion dollars in the data center developer iron or I'm not sure is it iron or Ern? You have a hold rating on that company. What are you kind of looking for for Nvidia to talk about with that partnership.
00:10:10
Speaker 4: I don't know if they'll specify it because in Vidia, because it generates so much free cash flow. As you may have noticed the last couple of weeks and months, they seem to be sprinkling the infield with two billion dollars for everybody like Oprah if they do highlight IRN know IRN is a neo cloud. And what they did with Iron it was pretty unique because first of all, they bought some capacity for Nvidia's own use, not to resell to customers. And then they also don't necessarily have a warrant, but they have a right to purchase more Iron stock and actually a significant amount, but it's tied to Nvidia in IRN delivering GPUs together, and I think they get the full payout once they've supplied Iron and its customers six hundred thousand of their GPUs.
00:11:00
Speaker 3: So it is a unique deal. It is a little bit different.
00:11:03
Speaker 4: But in Video, with all of its cash, it's been doing two things. It has been acting like a venture capitalist. The plant seats throughout the entire ecosystem. Remember my point that they're going beyond GPUs, and because they don't really have anything else to spend down, they're buying back.
00:11:18
Speaker 5: Some stock ed Lodo. You interviewed the Iron CEO on Monday. What was your take on the company's plans for this kind of partnership.
00:11:29
Speaker 8: Yeah, so, you know, co CEO, Daniel Roberts.
00:11:32
Speaker 7: You know, Iron is a neocloud that is largely dependent on video architecture. And what is so interesting is we're talking about the industry being in a position where demand is fast outpacing the industry's ability to supply. But you know, as Iron would tell it, the bottleneck is not in video being able to get the server designs or the GPUs into place. It is everything else. It is the utilities trying to get the instruction place to switch that data center online. Is the concrete, the manual labor, the tin can that needs to go around it. So it's very interesting to hear that experience through through fair experience of what it's like to try and build. And Neocloud's just a data center for away AI workloads, simple, right, but in video doesn't appear to be the hold up here.
00:12:20
Speaker 2: Let's bring back Jay Goldberg Seaport Research Partners analysts there, he again is the only analyst track by Bloomberg with a CEL rating on in video. Jay, I imagine that this report is not enough and I don't want to get in trouble with compliance or anything, but not enough for you to change your rating on in video.
00:12:40
Speaker 9: Nothing I've seen here sort of meaningfully alters the trajectory of how I think of in video's well you prospects.
00:12:48
Speaker 2: What is your bear what is your bear case on the company, and how do you see it? How do you see the stock falling so much in just the next twelve months.
00:12:56
Speaker 1: To just be clear, I've never told people to short in video.
00:12:58
Speaker 9: I'm always positioned I might take as an underperformed go back to when you interviewed me in May of last year. This is that was what I'm saying, And I think that's that's played out right. It is it's just underperformed. It's it's it's a good company, but it's hard for them to live up to, you know, five trillion dollar market cap expectations, right, and and we've seen that this year, like they've underperformed the socks and the SMH ETFs underperforming, they've they've gone up whatever, twenty thirty percent this here, I think, and lots of other names in the sector have gone up a lot more, and it's just a function of how big they are.
00:13:34
Speaker 1: And I think there's if you look at the results.
00:13:36
Speaker 9: I mean, one of the things that stood outs to me was they're increasing the dividend, which is a great signal, but it's a massive increase, and right, it's sort of signaling that they're running out of ways to grow faster, Like couldn't they've used that those funds in a different way.
00:13:52
Speaker 2: I just want to point out you've accelerating on the stock. One hundred and forty dollars is your price target, and that is a period of twelve months. Yes, okay, so you're not telling people to short the stock, but you don't. But you believe that the valuation should be one hundred and forty dollars per share, not two hundred and twenty dollars a share.
00:14:10
Speaker 9: Yeah, I'll obviously revisit that today, but certainly I think it's I think it's hard. It's hard to get excited about the stock when they are better opportunities out there.
00:14:19
Speaker 2: Paul, come on in. I imagine you disagree with Jay. There is it. Is it easy for you to get excited about in video still.
00:14:27
Speaker 4: Well, Jay had some very good points, and I obviously admire him and his work, but I would tell you that this is not a situation like Cisco was right before the Internet bubble popped, where it was trading in a couple hundred times earnings. If you believe, and I don't formally cover in Vidia, I cover the Neo clouds, But if you believe the consensus estimate for next FISC year now, this would be the fiscal year for Invidia that ends January thirty, first of twenty eight, which is really calendar twenty seven. Right now, we are trading at about half of the growth rate expected in that year. Well, Jenny had some very good points, and I obviously admire him and his work, but I would tell you that this is not a situation like Cisco was right before the Internet bubble popped, where it was trading at a couple hundred times earnings. If you believe, and I don't formally cover in Nvidia, I cover the Neo clouds, But if you believe the consensus estimate for next FISK year, now, this would be the fiscal year for Nvidia that ends January thirty first of twenty eight, which is really calendar twenty seven. Right now, we are trading at about half of the growth rate expected in that year, and we're trading at an S and P five hundred multiple. Actually I calculate about one click below that. And here's a company that really hasn't seen much multiple expanse over the years. It's all been in delivering the mail. And so my point is, I'm not necessarily disagreeing with the trend longer term, and it is hard to grow something this large, but we're treading it less than a market multiple. It's not actually expensive at all. If it was expensive, then I probably swing from bulla bear.
00:16:20
Speaker 5: But it's cheap, and I do want to bring you back in just going back to the China. The China story here, because we've seen this company's revenue, the revenue that they generate from China declining quarter after quarter. They're now continuing to exclude China revenue from their outlook for the second quarter. Does Nvidia need business from China at all?
00:16:47
Speaker 7: So there's the basics the numbers, which is the quarter did not include any Hoppa based systems to China. And the outlook for the current fiscal second quarter in financial year twenty seven, just as you said, does not reflect any China revenues or shipments to China. And what Jensen Huang in collect cress the CFO of said is that China from a starting pace of zero baseline zero is a fifty billion dollar opportunity. But if you if you again get out of the income statement, the story with China was a critically important end market for data center and as Jensen Wang would like to say, often fifty percent of the world's AI research or the researchers themselves are in China or from China. And so you know, if you buy Jensen Wang's mile high bigger picture view of what is going to happen with AI, the build out of AI training and transition to running of systems with inference to their mind, being a participant with China is critically important. And again you know in that interview Monday, and I'm sorry to go back to it, but you know the word he uses, he uses deliberate. The demand from Chinese tech companies is quote incredible. It's just they're not permitted by the Chinese government to do business within video.
00:18:03
Speaker 2: Yeah, certainly on the call a lot of questions and perhaps some more answers when it comes to the China business for Invidia. I want to bring back and remind everybody that was just Ad Ludlow. He's the tech Bloomberg Tech co host. You can catch him every day at Bloomberg TV at eleven am Wall Street time. Jay Goldberg Seaport Research Partners joins us too, along with Paul Meeks, head of Technology research at Freedom Capital Markets. Jay, I want to send a question over to you about the reporting structure and the reporting framework for in Video. I don't know if you had a chance to see this yet, Bart, this cut adds attention. It caught our attention when these numbers crossed at four twenty. Lin Dwan from our Bloomberg Technology team, she's a tech editor. She writes that the company's transitioning to this reporting framework that in the future, in Video will report for two separate market platforms, data center and edge computing. The company says the data center segment will include business with the hyperscalers and those building AI focused data centers. Edge computing will highlight the devices itself for PCs, game consoles, robotics, and the like. I know this is a little bit of a wonky sort of accounting question, but as an analyst like I know you want to see clearly defined areas when it comes to segment reporting. What's your what's your initial reaction to this new reporting structure.
00:19:20
Speaker 9: So at one level it makes sense. I have to think through it. But data center is really where people are focused on.
00:19:26
Speaker 2: It's kind of the number that only any It's the only number of people care about, right.
00:19:29
Speaker 9: It's the only number of people care about, with the exception of my my fourteen year old son, who only cares about the gaming GPUs.
00:19:35
Speaker 1: And it's going to be very sad, saddened by this news.
00:19:37
Speaker 2: Okay, they'll still get the gaming GPUs, they just won't be reported out separately.
00:19:42
Speaker 9: It's yeah, it's getting harder to get them, and they're not getting updated as fast as he likes.
00:19:46
Speaker 2: But well, that's interesting because that's the DNA of the company.
00:19:50
Speaker 1: That's the DNA of the company.
00:19:51
Speaker 8: Yeah, Ed, you're a gamer, come on in here.
00:19:54
Speaker 1: I'm a gamer.
00:19:55
Speaker 9: I'm a gamer, but your gamer two j I'm willing to forgive them on their piece of gaming GPU looks upgrades.
00:20:02
Speaker 2: And I know this is not the number that everybody's focused on. But since this is in videos d in videos DNA, and you're a gamer, what do you make of the gaming GPU issue? They're leaving it to someone else, Now what is it?
00:20:14
Speaker 8: I don't know that it's an issue.
00:20:16
Speaker 7: I mean, you know again, I'm just you know, in videos changing the reporting structure. I get that. So it's just not highlighting gaming. But gaming is not the growth engine. It's tiny for this company.
00:20:31
Speaker 8: It's it's tiny.
00:20:32
Speaker 7: But like again, maybe you just consider the GPU that goes into a gaming console or your you're home rigged gaming PC.
00:20:42
Speaker 8: It's still at the edge.
00:20:44
Speaker 7: What people say is at the edge you're running computation at a device local to you. And you know the story that jensmong has been telling on the road, right, I think CEES or GCC, whatever, whatever it is, whatever event, it's that the role of the GPU in the video games industry is now as much about the development of that video game, its ability to generate pixels, right, rather than animate That's the story for in video with gaming. The gaming industry is in Nvidia, GPUs and data centers still as much as it is about the GPU. And by the way, I got like this Frankenstein's Monster ASA. It's got an amazing and video GPU in it and an Intel processor and core. But like, so what man, It's all about the data center.
00:21:30
Speaker 2: It is all about the data centers. Hey, I'm speaking of the data centers. I want to bring Paul Meeks back in here. Ian King and Dina Bass, writing separately on the live blog, Ian talking about Hyperscalers still being the biggest providers of sales for that unit, but others a clouds industrial enterprise.
00:21:46
Speaker 1: Grew quicker quarter over quarter.
00:21:48
Speaker 2: It's a trend that investors will want to see accelerate to remove the concern that it's too dependent on customers who have their own components and an incentive to move away from Nvidia. Dina Bass weighing in and saying that at the same time time, the greater growth in non Hyperscaler customers does point to the idea that indeed, folks like Google and Amazon, while they remain huge in VideA buyers, are trying hard to expand the workloads they can put on their own chips and get more of their cloud customers to choose them over in Vidia GPU. You cover the neo clouds you understand what the hyperscalers want. What do you make of that sort of revenue diversity question? And the companies that have their own ability to do this pushing their customers into their own products.
00:22:31
Speaker 3: So I agree with the push and pull here.
00:22:35
Speaker 4: Obviously it's important that over time that enterprises go direct, not through hyperscalers, to broaden this market. I think it's a very positive trend.
00:22:48
Speaker 3: In Vidia, of course, will be a beneficiary.
00:22:51
Speaker 4: On the other hand, particularly this announcement the other day with Google. Yes, Google and Amazon in particular working very hard on their homegrown chips, and it looks to me like they don't want to just lower their own costs by using their own chips internally. But in Google's case, they're going to sell these chips on the merchant market, and so over time it could be a threat. Why because Google has very deep pockets, almost endless resources to compete when you have other companies that say they can compete over time with Nvidia, be careful because these guys have the best balance sheet there ever was. But Google probably can make a go of it, and I imagine Amazon will be on their heels
