NVDA是强力买入标的；Q2指引显示FY28营收增长约70%且供应承诺超$500B，证实需求加速并使该股被低估。

Nvidia's about to take over. I'm so excited to be here, ladies and gentlemen. Today, I'm representing Nvidia's digital twin, Jensen Wong.

this is going to most likely be the most important earnings that we've ever had for uh Nvidia.

Teimu Jensen, what are you thinking for this quarter for Nvidia? High level thoughts and then we can go a little further into the numbers. Uh, well, high level, I think that they're going to do uh very well. Um, can I jump into the numbers because I believe that they're going to essentially do triple digits in terms of their growth rate top line. Um, my estimate was a little bit conservative even though they're higher than Wall Street's numbers, but 93 and a half billion. I actually think that they could come in probably over a billion higher than that. Um, but that would be in a really good bull case. Uh, so my sort of base case is 93 and a half billion. That would be 100% growth rate. And then on a non-GAAP EPS base, and I already know the numbers because, you know, I created the company. So, uh, we're going to see $2.16 of non-GAAP EPS, plus because of my amazing investments in SpaceX, we're also going to see an additional $10 billion worth of gains, which could end up seeing, you know, $257 worth of gap EPS due to the market marktomarket gains that uh, you know, Colette did for me. So, so she's going to mark some of those SpaceX gains in this quarter.

I forgot about the SpaceX gains. Do you think the street's going to care about that or it's like it's baked in? They know there's EPS growth from that. Uh well, I mean, there will be EPS growth. I believe $10 billion worth.

No, but if they beat EPS by like 10% and it's because of the SpaceX stake. Do you think the street's going

No, no, no. I I think non-GAAP uh Wall Street's expecting $29. I expect $2.16. Even without the marktomarket gains, I think that they'll beat without it.

Please make sure Nvidia goes up after earnings, otherwise I'm unsubscribing. I think all of us have been waiting for this stock to finally make a move. And you know, the implied volatility on the name is 5%. It's the lowest it's been in two years. Usually, it's 8 to 10%. So, the fact that there's not that much volatility going into the sprint means people have kind of given up on expecting a big move, which means maybe this is the quarter where we where we get the big move.

uh results wise, I think the market now is actually expecting 100% year-over-year growth. um a couple of weeks, maybe months from now. It was much lower than that, but the market is already expecting triple digits now. Maybe they're putting the bar higher, so they are not so far behind, but I think Nvidia delivers quite easily on on a double beat this quarter. Then it's all about the guidance for next quarter. Tenra or Jose, the next quarter is what closer to 90% year-over-year growth or something close to it. I think they can deliver on that as well.

the bigger story will be on all of the commitments, all of the deals. Uh how how are they thinking about it? Maybe some questions on the whole financing agreements that and the announcement that we've seen, how are they choosing it, how is that structured, etc., etc.

But all in all, I expect Jensen to have another call where he just tries to explain the what the heck's going on right now. There is no scenario in my mind where Nvidia shows weakness. Nvidia has visibility into 90 plus% of all the the companies that are using AI and GPUs and whatever in the world which means they know what companies will be doing two years down the line or trying to do two years down the line and and so how how can they go and be behind it just it it doesn't happen. It doesn't happen. So, I expect a full beat. It the stock itself has been underperforming this year. I think the last time we covered this, the stock was at the exact same price or or close to it. So, it's time for this one to get to get a move on.

Jose, you have uh owned Nvidia for a long time. Your thoughts on on this quarter?

Um, so like Neil was mentioning, I do think one of the these are two topics that I think are going to be big discussions, right, in today's starting call. It doesn't mean I think they're important, but I think it's what the market is going to be focusing on. First is kind of the uh jalapeno chip, right? We kind of saw that with OpenAI yesterday. Um, and it's it's become very popular. I think it's social exploded. I saw a lot of topics and discussions about this. It just continues to add that sentiment that Nvidia, you can't hold your market share forever. Look, AI chips, AI is making AI chip development easier, which I think is true, but I think the the moat is changing a little bit. It's no longer about chip design. It's about the supply chain. Who who has the ability to be able to lock in 60% whatever x percentage of TSMC's manufacturing? who can hold up x percentage of of the HBM market or the rest of the manufacturing space. Sure, you can design a chip, but can you go to all these kind of supply chain players and make them promise you that they're going to give you a piece of that supply chain for a chimp that they don't know is going to work while you have Papa Papa Jensen and Granny Jensen here kind of giving you plenty and plenty of money for for that for that allocation.

The second thing I really would want to hear about is just like Neil was talking about the um commitments, right? And see, I was listening Neil um I'm giving you your uh but it's last quarter I think we were closer to like 120 $130 billion in commitments and I don't even think that was including kind of uh cloud plays and it wasn't including investments either. If we see that number kind of go to 200 billion, as um as an investor, I would want to know that visibility. You've given us $1 trillion of visibility between 2025 and 2027. But if now you're increasing your supply chain commitments dramatically, I think it's only fair you tell investors why. I already know you have it. Just give me a number. Tell me a number till 2028. Tell me $1.7 trillion to 2028. I think that's what could wake up the stock and say, "Look, Nvidia definitely has this and this is why they're making the commitments because now you're telling me that you've increased your commitments dramatically and you're still sitting at that $1 trillion till 2027." It just doesn't make sense.

I think I think I think that's a big point, Jose, because when Jensen announced he has 1 trillion in visibility going into uh 2027 calendar year end at GTC, you and I were there together. the stock popped every and then like two minutes later the market was like wait a second we already knew because the whisper number was like 950 so it's only 50 billion more so you're right I don't know if they make that announcement here but they're going to have to say that Vera Rubin and everything they're seeing in terms of progress with the hypers scales and ROI is got to push that one trillion to like one and a half at least to get investors a bit more excited about the multiple expansion

so I I would say um that commitment news and hopefully any visibility that's what I'm really looking at I I I I don't know why I I stopped caring about the next quarter and like I really don't look at it. It's just I know Nvidia is going to crush it, right? So, it's like uh just tell me what's what are the years I don't I don't see coming right now.

Um but yeah, I mean the other thing that I think will be curious is they did break down their data center revenue segment to kind of the meta uh the hyperscalers and the um neocloud players or whatever they forget neoclouds. um enterprise and so on. Last quarter that was almost a one-toone basis. How has that shifted this quarter? Right? They mentioned that the NeoCloud players were going to grow dramatically. We heard from SpaceX that they're going to do exclusive Nvidia. Are we going to see that shift? And does that eliminate that fear of ASIC players, right? One of the fears is that your Googles, your Amazons, everybody's designing their own AI chips, but now Nvidia is giving you a clear number that says, "Sure, they're designing their own chips, but the ones that are really buying the chips are these guys right here. The ones that aren't really going out and building those chips right now." So, I want to just see that number as well.

we are likely to be the most free cash flow generative company like in terms of a single quarter than any other business before. So, Apple I think did 51.5 billion dollars of free cash flow on December of 2025. um that is likely to be beat by Nvidia if we end up keeping up the high free cash flow margin that we've end up seeing.

I also want to know what they're doing in terms of the buybacks, how much they've actually purchased this quarter because we've not seen uh a large appreciation over the, you know, 3month period or at least there's been opportunities to to buy in at really cheap prices. That is the same quarter where we ended up seeing extremely low prices in Nebus and Coreweave and we really saw a dip in a lot of stocks and uh Nvidia was down to I think like 170 or so. They have I think it was $39 billion plus an additional 80 billion on top of that of uh stock purch stock repurchase authorization. So if they decided to buy back a lot of stock, that might also help the EPS as well if that comes in much much lower than Wall Street expects.

um 68% of you say Nvidia is going to outperform so obviously there's a lot of excitement for Nvidia having the best numbers which I think they will once again it's going to be a question of if the street is willing to reward that do any of you feel their gross margins will be under pressure a little bit or should we expect it to be at 75?

The only reason would be you're starting that Vera Rubin um ramp and like we saw kind of like the the uh the hopper to Blackwell there was a bit of a hit. Um I actually don't remember what was the guidance tenant. Do you remember was the guidance that they gave us? Um I would just pretty much if they gave us that number I think we're we're pretty in line with there. The only thing I would also want to look into is with the price increase that we saw or heard about um was that a pricing power move or was that a memory infle in in in um inflation cost move? I I think it's more the latter than anything. Um but how does that shift the um how does this shift gross margins going forward?

They guided for 75% margins. Um, and operating expenses I think was like 8.3 billion. But um, so yeah, gross margins are expected to come in right where they were before and they didn't come out and give us any preliminary numbers or any changes. So I expect them to be high. Some people talked about memory uh prices putting a lot of pressure on the chips and that their recent 15% bump up in Blackwell and Vera Rubin um prices being an indicator that they might end up coming in shy this quarter on gross margin. But this is not the first quarter where we ended up seeing memory prices increase and margins have only improved over the last, you know, 12 months or so where we've kind of gone from that 72 upwards of 75% climbing up. Um, where were the pressures back then? So, I'm not sure what to expect because it's not like this was a one quarter hit. Memory has been a a little bit of a drag for a while for them.

Rubin um prices being an indicator that they might end up coming in shy this quarter on gross margin. But this is not the first quarter where we ended up seeing memory prices increase and margins have only improved over the last, you know, 12 months or so where we've kind of gone from that 72 upwards of 75% climbing up. Um, where were the pressures back then? So, I'm not sure what to expect because it's not like this was a one quarter hit. Memory has been a a little bit of a drag for a while for them.

I would say for Nvidia too, I mean a lot of these manufactur for the memory, they've been locking in supply. So, it's it's if there is any memory changes, it's not going to affect this quarter or next quarter for them. Like, they already knew what they were going to be paying for and to make these servers, the memory was procure maybe a quarter ago or two quarters ago. So, it's not like that's going to impact margins today. And I think that's why they if they are increasing the prices because of memory, they're proactively doing it because they've heard that, hey, look, whoever is selling me G uh memory, they're telling me that next year or a little bit further than that. That's when we're going to see a higher increase in memory prices for us. So, for us to proactively keep that margins, we're also going to increase uh pricing. But I don't think we're going to see um they should know the pri if if you're building this type hundred billion dollars in revenue per quarter. You know what you're paying for your memory many quarters ago. It's not going to be a surprise for you. So it's not going to be a surprise to margins.

couple of people saying Bur disclosed that he bought Nvidia calls. That is true. They are a hedge to his Nvidia short. So uh that is why you're seeing that headline right there that he bought calls.

Do you think that he ends up adding to his shorts whenever he sees streams like this? These are the opposite purchasers of his shorts. I I I just think he he once again he's looking for big short 2.0 and AI is if AI is the second biggest short in his career, Nvidia's got to be at the center of it, right? So, but if he's wrong and because Nvidia has been taking a bit of a dump over the past few days going into

Jose, do you think Jensen's going to have to calm people down again around circular financing, AI bubble, all of this hype, it's getting people too excited, so you know that skepticism has to be calmed down. And then do you think he also also calm down investors around the political backlash that we're seeing from red and blue states to uh create the progress for more data centers?

Um yes to both. first to the the first one of like the circular financing. Um I think the only way he can really calm investors is by giving that visibility further on or or the higher visibility. It's like look we're doing this yes our obligations have increased by an extra 40 50 billion this quarter but that's because our visibility in revenue has increased by 500 billion or or or more right so I think that's the only way he can really calm that side. Unfortunately, on the political side, I don't think there's anything he can say outside that just AI needs to continue to innovate in the United States and it's for national security and it can it's going to be a huge growth in GDP. But I I don't think it's it's enough. I there's nothing he can show that's enough that can eliminate the political fear. And I think that's going to be one of the biggest um I now every quarter or in every segments or sector or or every few quarters there's a new bearish thesis. I think this is the bearish thesis that unfortunately Jensen won't be able to give uh a good answer to to com investors.

Tanner, how do you think Jensen should uh navigate this earnings call? Yeah, I mean he's definitely going to talk about circular financing. Um, I think that there's been a couple things that I can almost guarantee, well, I'm not going to guarantee, but like I really think he's going to touch on SpaceX. He's talked extremely positively about the companies that have backed him and have purchased his products. So, um, I think he's going to say a lot of good things about SpaceX this quarter because of that exclusivity comment that he made. I think they're definitely going to bring up what Coree said about the amp years lasting over 9 years. Um, these are going to be some of the highlights. He talks about circular financing. Um, the Wall Street deal. This is one that they're going to talk about in terms of creating uh or or talking about Nvidia's chips being so reliable that they can become uh an investable asset that their performance um you know could be benchmarked probably taking on more risk. Like I I think a lot of those comments there's nothing but concerns on Nvidia. I don't think you can talk to any real investor and say the um multiple for what you're paying on this company right now putting up 100% growth rates is expensive. It's just not true. But it could be expensive if you believe that things are going to go very wrong for them. And that's all they're going to be doing today is just dodging those types of bullets.

there was actually a recent upgrade from Raymond James to $352 from Nvidia. uh those estimates continue to climb higher.

do you think Neotron and Poolside acquisition will make a difference as Nvidia pushes into open weight models? I think the reason they're pushing into openweight models is because all of their biggest customers, Open Eye, Anthropic, Google, Amazon are building their own chips and pushing into Nvidia space, competing in the chip space, which is smart for them because they don't want to keep paying the NVIDIA tax, but obviously Jensen has to fight back. So Neil, do you think there will be a heavy emphasis on the call around the the benefits of open source?

um 100% is going to talk about open way models, um open source, Neotron, the recent Aquire. I don't think they'll talk that much about it. Maybe maybe he'll he'll name core loop side but yeah the drop in token prices he's going to say look drop in token prices it does not mean that these companies will generate less revenue just means that now we have GPUs that you can charge more per hour they'll just finish a job way quicker and so your GPU frees up quicker which means more customers can use that same GPU and you'll generate way more revenue it's it's all about how many the cost of what million tokens or something like that that uh the equation that he always uses um he will again say that we are the best when you look at total cost of ownership um nobody can can go against us and so it's going to be again the same story that that he've been saying time and time again whether the analysts and the market will listen this time is is up to them but from his standpoint I mean I've talked about this with Jose a couple of days ago with regards to closed models again going against the the open source, open weight models for Nvidia. It's it's going to be a win-win regardless. So, he's just again positioning itself himself as okay, you know what, I'm going to win regardless of what happens. And it seems like we are going to see a move towards way more open way model than uh than the other direction. That doesn't mean that Anthrop Anthropic sorry and open AAI are going to go out of business, but it does mean that Nvidia make sure that they are going to win.

I I I also wonder how he's going to frame this open way. Is it more of a defensive move against the AI companies or is it more a defensive way, a defensive play against kind of China? Because right now the openweight models are being built uh are mainly in China. So Jensen is a very very he talks about a very paranoid person. So I wonder if his paranoia is saying if if AI Chinese players keep developing Chinese AI players keep developing these models eventually aren't they going to go to Huawei or or or or whoever builds chips in in China and start getting their chips manufact um like made more efficient on Chinese infrastructure. So Jensen is like I don't want that to happen. let's create a strong open AI model world here in the United States because then we can assure that the chips that we're going to be making are going to be used for this openweight solution. So I wonder if it's more of a uh regional issue opposed to a defensiveness against the frontier plays or maybe even a little bit of both.

How many people in the chat you could just put what you think is is thinks that it's going to be flat. How many of us think it's it's going to be like a 23% move like it's not going to be that substantial? I think there's there's a decent chance 209 to 212 and it's just there. The market's not happy. The market's not upset and it continues to do what it's been doing for the past couple of months at this point. You think it's just going to trade sideways forever?

No, but I I wonder if like I wonder if the results are amazing but not enough but it's also not enough to to take it down because you know Nebia's core we have all these semis are going to be contingent on what Nvidia does today. And if it does consolidate and go sideways it wouldn't surprise me. I don't expect it to happen but I don't think it would be surprising if it does.

Yeah. But I think that's more based on um like the history of the stock is getting to you, you know, like uh similarly to Coreweave's earnings, it had dumped every single quarter before that everyone was residing to it dumping again. Then it was up 19%. So, and I'm not saying that this is the breakout quarter, but at some point I I just read the report from Raymond James. They talked about by 2027 they're going to have a lower PE than um the S&P 500. How long can that go on for if the growth continues to stay up? How much of their outstanding shares will they be able to buy back? What does that dividend growth get you? You know, they just 25x it. Are they going to end up uh able to 5x it again? I mean, they've got the money to do so. So I I like at some point I think that the valuation has to bottom out. Maybe it's a year from now, but if the growth is expected to stay high. Um yeah, I don't know. I'm I'm not counting them out.

To me, the only way this stock moves up today is if we get a strong visibility number. Not not just even a strong guy. Like even if guidance is crazy, I think of visibility to be on the number they given us that 1 trillion. I think that's really the only thing that can say, "Oh, you know what? Maybe we aren't crazy in AI build right now."

Neil, you think it can go flat or something? If it if it does drop by the expected what is it, 5% or so, we are I just checked with my crystal ball, we are going to go quite close to the 200 day moving average. So, I don't know if if that's already pre-planned by by the market, but whatever.

So, I don't know. I bought an extra share today as always. So, hopefully it goes up. So, I'm up on that extra share. It's a big one today with Nvidia Q2 earnings.

So, we do have a bunch of these other companies that we will take a look at before we get to Nvidia.

Again, Nvidia's in about 20 minutes. But Octa and Salesforce going green.

I also think it's good if the software names get some love because at the end of the day you know Nvidia a is you know trying to make sure that the the GPUs that they sell to train these models are helping the software companies. Ironically the software companies create a lot of demand for those um the Frontier Labs. So, it would help Nvidia if AI is helping software, not destroying software.

If Nvidia were to actually fall, I would be a bit skeptical on why given the software companies are growing because of AI. Nvidia is supplying the GPUs to make AI a thing. So, I don't know. I think these software companies are giving us a pretty good read into what Nvidia should do.

Nvidia right there 21038 now uh after hours. Does anyone here think they could actually miss on earnings by any chance? And if you don't think they can miss, Tanner, do you think you know Streets Guiding for the Whisper number is 97%. Do you think Jensen just goes that extra mile and gets them the triple digit growth year-over-year?

Yeah, I mean they're they're they're going to see 100%. Well, I mean, I can't say that. We we don't know 100%. Um, but I believe that they would push that over to 100% if they were close. I mean, it's just a really good headline number. Um, that could get Wall Street or ALGO excited. But for real, I it honestly could be way above that. It might even be closer to about 106%. Um, seeing roughly about 94 95 billion, which many analysts have also reported on.

As long as this continues, then I think Nvidia should be able to show some strong growth as well.

Okay, Nvidia numbers should be out. We are down 1% on the immediate reaction. 20760. I don't have the numbers yet. Let's see if that reverses at all.

Nvidia 211. 96.2 billion in revenue. Cleared my number. Absolutely cleared it. Adjusted gross margin 75%. Net income 59 billion. I don't know why it's down. There must be some reason. Guidance destroyed. Revenues up. Net income big, big beat. What is the street not lagging? How is this not 220 right now? 96.2 billion in revenue. Nvidia grew 106% year-over-year.

EPS is up 111%. 22 versus 209 for EPS. 96.2 billion. That's a $4 billion beat versus 92. Stocks down 2%. Of course, why wouldn't it be down?

The growth here is going to be fantastic. Even the guides way higher. Like what what does that guy rep guide represent year-over-year growth on? Let me do the math on that real quick. Data center revenue was 89 billion versus 85 billion. I'm trying to find the guide right now and get the press release. Uh stocks at 205.

So, revenue up 106% year-over-year. Adjusted EPS up 120% year-over-year. Data center up 117% year-over-year. Adjusted gross margins actually are up 250 basis points year-over-year at 75%. Q3 revenue guide. The street thought they would do 104 billion. They are guiding for 108 billion which you know that means they're probably going to do 110 to 112 billion which means we're going to get two quarters in a row I believe of around that triple digit growth.

Adjusted gross margin that's down next quarter 74 versus 75. Adjusted opex at 9 billion. Um operating income for this quarter was 64 billion. That's up 124% year-over-year. They I believe bought back with the dividends 26 billion in Q2. They have 99 billion left on their buyback plan.

Jensen says AI has reached its inflection point. It's doing useful work. Its tokens are productive and profitable. Now compute is revenue. The AI infrastructure buildout is at full steam. Bar Rubin now in full production was built to power exactly this moment.

They've also broken down the revenue for the first time or for the second time because they started this last quarter between hyperscalers and everybody else. So out of that uh revenue that they got 96 billion 48 billion of it was hyperscalers Amazon Meta Google etc. The street thought it would be 43 uh the second version is called edge computing this is basically your neoclouds 7.2 billion versus 6.7 that the street thought it would do. So they beat on hyperscaler and edge computing. Obviously the majority of their revenue is still concentrated within the hyperscalers.

So it's 90% growth for next quarter approximately. Exactly. And and it's likely to be closer to 95 if they end up putting up or maybe even a little bit higher.

This is uh quite incredible in terms of the stock not really getting the move. you can see shares are dropping uh roughly 2%. Could be because of two reasons. First, the Q3 revenue guidance coming in at 108 billion, which is higher than what the street anticipated, but lower than what the buy side numbers were around $ 109 billion. This revenue guide again does not include anything from China despite all the back and forth headlines for data center revenue. In the quarter that just passed in Q2, data center revenue came in at $89 billion, slightly higher than the 86 the street anticipated. And then a growing concern was gross margins for the quarter. Gross margins came in at 75% exactly in line with what the street wanted, but it's the Q3 gross margin guide that is light at 74%. Uh again, we know that they had promised mid70s and that there was going to be some tapering off because of memory costs. And so now we're starting to see that right now. if it's actually getting sold off on a 1% gross margin miss.

I don't think that's the reason for the selloff. Um their commitment numbers are out. Um commitments were increased from 119 billion last quarter to 279 billion this quarter.

Yes. Nvidia making commitments to like its supply chain players to its neocloud players to whatever to allocate whatever type of resource right either being land power the land power that it's doing uh the supply from memory the supply from TSMC so last quarter that was 120 billion this quarter is 279 billion a bear I think anybody can say hey that's getting crazy numbers right 279 billion like maybe when you were at 100 billion your free cash flow and all this would be able to cover you just fine, but now you're going beyond that in your in what you typically have in your balance sheet. I think it's extremely bullish because if they're doing this, it means that they have so much visibility. And this is why I go back to where if you're giving me this type of growth, if you're telling me in the in the quarter you grew your your commitments by $150 billion, you have to show me proof as an investor. I think it's it's right to to for investors to ask to tell me what visibility that you see that is making you do these types of commitments. But 280 billion I mean you're talking about the sizes of certain semiconductor companies in just their commitments right now.

Yeah. But the commitments are for 27 28 29 2030 2031 2032 and thereafter like the the 120 is of course for the remainder of 27. Then we have 100 for 28 and then the number does come down from there. It makes sense, right? Because they they've made some huge announcement with with memory players, with some supply chain players. So I'm not that surprised that the number is up quite a lot. It's just them securing the upcoming years. Um it's a big jump, but I'm I'm not that surprised. like pri say primarily related to the procurement of memory.

It's that where it's like and I agree with Neil. It's not bearish in my opinion. Similar to Neil, I'm pretty sure Tanner, everybody. It's like this is what you need to make sure you don't lose that supply to somebody else. If everything is a is a bottleneck right now, you don't want to lose that supply. you need to make these promises to those players. But then the street wants to know, hey, if you're doing this, and I'm happy that this is the first time they actually broke down that commitment in this press release. I think usually they share it only in the 10Q report and not even to this extent. Um, but we we now see kind of 2026, 2027 as Neil was just mentioning.

I mean, if they're committing this much money, even if it's over 10 years, the bull case is that they have so much vis visibility into demand. The bare case is they're going to be spending more money. Their margins are going to go down a little bit. Like, is that a reasonable enough bear case?

Yeah. I mean, I feel like we knew this with the Wall Street deal with the Ohio data centers that they were increasing their overall commitments. Um, I mean, for me, it's business as usual. They also returned a record. So, so Free Cash Flow didn't end up beating um Apple's 51 billion because they returned a record $26 billion to shareholders. Uh which pretty wild. A lot of that in share repurchases, some of that obviously in dividends.

the EPS came in so much higher than I would have expected. The gross margin came in exactly where we thought it was going to be. Maybe not for for next quarter, but there's always um potential inflections in margins whenever we have uh whenever we're going into new architectures like that lady said uh on CNBC as well that none of this is inclusive of China revenue at all. And we also are seeing China loosening their stance and are allowing tens of thousands of GPUs to be sold. Not a lot. These are not big, but the fact that they're not being included at all and there's likely to be hundreds of millions, if not billions of dollars worth of GPUs uh sold to China is really important.

Um I'll have to listen to the whole call. Like I'm not going to say whether that's bullish or bearish, but this is beyond my expectations so far and the stock is down on this news based on commitments that we already knew was going to be a part of their game at least for the time being.

Yeah, I think Tener the only thing is more like the size of the commitments and I think most bulls understand it, right, Tener? I think we we we understand that we know this is needed to continue to see this growth rate. Uh but I I also don't think it's it's impo it's it's bad for market to be questioning because I'm looking at the numbers some more in that press release and it was it was for from here to 20329 billion on supply commitment 29 on cloud service agreement a a total of actually 366 billion um from now till 20 uh till 2032 and thereafter just on supply commitments but then they also have other types of commitments they have the AI AI cloud back stops which is different from the AI the cloud service agreements that they had in the previous one but in forms of AI cloud agreements um and data center leases that's an extra 56 billion to that 366 billion that they have and then then they have guarantees on those kind of land power and shells and on that already they have about 108 billion. So you're talking about in all these commitments and if you scroll down a bit uh on it there's like the 366 here then if you scroll down a little bit they have the the next page they have 56 billion here on other commitments and then be uh you just passed it the one on the top then they have 108 billion here um one under there yeah 108 so you're you're talking about commitments now being close to $500 billion right so it's I I agree that and Diddian needs to do this. I I I truly do. But it's I think the market is right to question, do you have the funds to be able to do this? Is this AI cycle going to last two years for you to be able um to to make it? Um again, I think so.

but it has to there's no way Nvidia can say we're committing upwards of $300 billion if they don't. To your point, Jose, you said I think this is the most important thing you you said in the show out of all of us today. The the visibility on that revenue number has to exceed a trillion at this point if they're already commit.

It's kind of funny. The hyperscalers sold off this year initially because the capex was going from them to the capex beneficiaries micron debas etc. Now it's like Nvidia is getting sold off potentially right now because they have all these commitments to those same companies but they were the ones that were getting all the the capex from the hyperscalers. So I would imagine you believe Jose they have to give you that visibility of revenue number otherwise this is just too much of a commitment for a trillion dollars over the next couple years that the street's not going to be happy about.

Yeah. I mean I I agree they need to give me that if we see I I'm actually debating right now if I should buy some stocks right now because if they're seeing this type of if they're making these types of commitments even if they don't tell me the number I already I it's it's going to be some crazy number that they have in visibility like it's

No. No. Nvidia. Like it's making me so bullish in this company that the type of com like if the street is punishing you because you're preparing yourself for the type of of visibility that you have. I I think it's I mean I like it. I I'm trying to think why the market hates it, but I I like it right now.

They did say here when under the guarantees that the the commitments like the each generation of Nvidia infrastructure deployed at port spy could represent approximately 1.5 million Nvidia GPUs or approximately 150 to 200 billion in Nvidia revenue and the sites over the next 20 years can support multiple infrastructure upgrade cycle. So they they give us some clear numbers here.

this earnings call just got a lot more interesting. Because now that these commitments are in the pushing three400 billion range, it's, you know, this is the obvious bare case that someone's going to say, why is Nvidia willing to spend a couple hundred billion dollars on all their suppliers? Because they're financing the buildout of their own chips, right? like for for for them to sell these chips.

Because now that these commitments are in the pushing three400 billion range, it's, you know, this is the obvious bare case that someone's going to say, why is Nvidia willing to spend a couple hundred billion dollars on all their suppliers? Because they're financing the buildout of their own chips, right? like for for for them to sell these chips. Everyone in the ecosystem, Micron, SKH Highix, the Neoclouds, packaging companies, they have to also be able to support because there's all these other, you know, nuts and bolts, photonics, memory, etc. that go into the chips. So, if if if some of these companies are insolvent and they don't have the demand from Nvidia, which for many of them is their biggest customer, Nvidia can't sell any chips. I I think that's going to be the bare case here. I think there's an easy answer to that. As Jose said, that means they have a lot of visibility into demand. But Jensen's gonna have to really explain that on the call today.

So, if if if some of these companies are insolvent and they don't have the demand from Nvidia, which for many of them is their biggest customer, Nvidia can't sell any chips. I I think that's going to be the bare case here. I think there's an easy answer to that. As Jose said, that means they have a lot of visibility into demand. But Jensen's gonna have to really explain that on the call today. Nvidia right there still down 2% after hours.

At the end of the day, he will just say, "Look, our new chips, if you have the same megawws, you will be able to generate more revenue. You won't need, not everyone needs a gigawatt scale data center. If you have 100 megawws, you previously made X dollars. With our new chips, you will be able to make X at the power of I don't know what's the the jump here between Blackwell and and Vera Rubin, but it's quite it's quite big and and he's just going to repeat it again and again and again. Um I I honestly if I were looking to to expose myself to even more, I I I wouldn't mind buying it more now.

Um I I honestly if I were looking to to expose myself to even more, I I I wouldn't mind buying it more now. The software name is not budging. This is a this is a very specific Nvidia reason for why we're down right now.

and what's interesting is all of the neo clouds are green as well so something doesn't make any it doesn't make sense like if Nvidia is down I would have assumed the memor the Neocloud names would be down as well, but they're not. The semis are either flat or greenish.

Well, well, I don't think No, I think that goes to point to Jose's point being right is that the AI demand is definitely there. Chips are, you know, selling faster than ever. Um, and actually you can see them break down hyperscaler versus their other AI clouds. So, so there's more growth headed towards the Neocloud names than the hyperscalers.

And then on top of that, Nvidia is willing to take on a lot of the risk. So, just individually, this is not a statement about the AI trade. This is an this is a statement about what's the risk level for Nvidia, but I think the risk level has always been higher from the purchasers of the chips, the ones with the, you know, mega capex bills. Um, but yeah, they're just starting their their bill, I guess.

To to be fair, the stock is down 2%. I mean, it's 1.2%. Like, it's not it's not about the stock reaction that that's that that that's that compelling to be upset or excited about. It's more so about this commitment number that now creates a new category of concern which again I don't think it's an actual concern but it creates a hurdle that the company will have to be able to articulate well enough to explain um for investors to deal with. I mean it's already the biggest company in the world already majority of the S&P 500. This is just a new thing that they're going to have to break down. I would imagine once again it means they have a lot of visibility into demand. They've already hinted at that in the CFO commentary. So on the call they would have to continue hinting at I mean, if the stock was up 5% after hours, I think the interpretation would be like, "Oh, they're spending all this money because they're going to make all this money." But because the stock's down a little bit, it might be seen as a negative.

People are saying, "When is Jensen talking?" I'm talking right now. That being said, um the call is at 5. Jensen doesn't talk for a while. He joins during Q&A. So, people always put up a bunch of comments during that. He uh does not do the initial numbers or the CFO commentary. He'll come in for Q&A. So you'll see his actual words around 5:30. So stick around for that.

Nvidia CFO says we partner with an extensive network of suppliers to secure critical components needed to meet demand for next several years maximum exposure under land power shell guaranteed 3.5 billion. Yeah, I think that the entire call is going to be about them saying the reason we're doing this $500 billion financing thing, the reason we have all these commitments is because I think Jensen's going to be like what are you guys talking about this AI this is going to the moon and we have all the money in the world. We have the best margins. We have to spend this money and we just bought back 20 billion of stock. Like you guys can't really be mad at us. Uh but we're going to keep proliferating the buildout of AI. I think he's going to have a very optimistic tone here and you might see the stock potentially flip on that narrative.

That's a win. It didn't hit that magic uh similar growth rates quarteron quarter adjusting for China, but it is I think that is a win as far as the guidance. You know, that was 3% upside last quarter. they guided 4.4% upside. So that magic number of 109, we did miss that. The stock initially, that initial trade-off was related to the gross margins. That one surprised me. Uh I was expecting more of a gradual decline of the next year, not seeing that step down in the October quarter. But at the end of the day, we're splitting hairs here. I mean, this is we have the the the metrics they miss fractionally. That's what we do. But I mean, the the aggregate of this this story is still very much intact. They're beating at a similar rate that they have over the past year. They're raising at a similar rate despite higher and higher numbers. And you have some of these big drivers still. Let's just look at SpaceX for example. They want to do eight gawatt hours next year with the data center. I mean, just that alone, just SpaceX alone can become a 3% customer last year to a 10%. That will take the hyperscaler growth rate from the streets looking for 38% next year to the low 60%. There's still some meaningful drivers. So Mike, I think that yes, we didn't check all the boxes perfectly, but the kind of the story is very much intact.

So, so Cody, I look at, you know, the kind of the setup here. Obviously, investors, the buy side has a little bit of skepticism or they haven't felt motivated to really rush and and urgently pay up for Nvidia shares for a little while now. On the sell side, 63 out of 67 analysts have buy ratings. The consensus price target is for a 50% gain from here. The consensus price target implies a $7.5 trillion valuation for Nvidia. So there's enough optimism as people who are looking at projecting the fundamentals. What do you think clients need to understand if in fact this stock is going to get unlocked to the upside?

Well, I think you've just got to factor in the continued growth of the name and I think the stock will eventually follow. uh when you've got this kind of of growth rate even on the hundred billion dollars of revenue uh continuing to double their hyperscaler in their customer base uh year-over-year and as Jean mentioned uh there continues to be a lot of new opportunities out there uh in the future and so I think if you put up the numbers the stock will eventually follow uh

we have two new disclosures from the press releases number one is one direct customer represented 16% almost a fifth of the revenue. Last quarter they told us their top three direct customers 21% of total revenue 17 and 16%.

No, I mean it uh Jean just said that if let's say SpaceX magically go from 2 GW to 8 gawatt next year, they will become a a what 10% customer. So, so, so clearly customers with a lot of capital can become much much bigger and then you have less concentration because suddenly you have another gigantic customer that pops out of nowhere which is SpaceX AI.

Neil, did they disclose their China revenue last quarter? Do you know? Yeah, it was zero. Oh, really? Close. I don't think this is true, by the way. That that I just read a report saying it was less than 1%. Yeah, I think China I think China revenue is a bit different. Sometimes they still include consumer products which can still be sold on Nvidia uh on China. I think the big news here was that there was a 1% of data center less than 1% data center of hoppers that did go to China. So that's the first time in a while that we actually have seen hoppers go make it to China. It wasn't it wasn't anything crazy, but that was kind of the big uh one of the new things here.

Yeah, there was a report halfway through July, so it was only two weeks where they were able to sell into China, Q3, I think, will be a very uh different story. And if you get an entire quarter of of sales into China, including a step up or an easing from China halfway through in August that we just saw a report 10 days ago, um that could be really good for uh China or for Nvidia considering they're not guiding for any revenue, zero.

Okay, so you have uh one customer doing about 16% of the revenue kind of similar to the concentration last quarter. Uh 44% of the first half total revenue was from three direct customers. Nvidia put in the press release. They also incurred a $400 million charge in the first half of the year from excess H200 inventory.

Going to China in Nvidia incurred a 400 million charge in excess H200 inventory in the first half. Yeah, that was probably inventory that was supposed to go to China that didn't make it, right? Most likely. Okay. So, it doesn't seem like that's the biggest deal. Wasn't there also like a big H20 charge last last quarter? Yeah, like 5 billion or 5.4 billion, something like that.

Yeah, I think a lot of people were upset about it wasn't accounts receivable last quarter, but it was uh a different accounting metric that kind of exploded. I think commitments is the big one that people are talking about this time, but that's something Jensen will probably talk about. Nvidia says it will be shipping both Blackwell and Reuben systems in the future and is currently experiencing certain supply constraints in their SEC filing. Nvidia said that in August it entered into with large capital providers. This is the big $500 billion deal to create independent financing platforms to mobilize third party capital

Gene Monster said it really well like we're splitting hairs on gross margin next quarter coming down 1%. Like most of this stuff we are looking at every minute detail. The two biggest details here is this is a company growing at 106% year-over-year. Guidance is is honestly very similar to that level of growth at if you know if you include what they're likely to actually come in at. And then you also have their increased commitments. Those are the two stories. Extremely fast growing and potentially taking on a little bit more risk or believing that this buildout is going to continue for a while. You could honestly look at that as a bullish sign thinking that they believe that they have visibility into multiple years of customers of holding up 100 plus billion dollar revenue quarters.

Like I I think the market's being crazy right now. Not only like the 2% being down or you know right now8% being down is fine, but it's off of the back of almost like an 8 day streak of being red. That's the part that's like frustrating me is like we are already down. Um

I think as as this person says in the chat, nobody cares if the stock doesn't go up. I mean they're right like all these little things were going into the nitty-gritty. No one cares. Everyone's like the stock should go up. But you have to analyze this stuff as Tanner says to try to get a little bit more deeper into why it's not going up. And you know, at the end of the day, it's been in a downtrend for the past week as you said, and it might continue that downtrend until Jensen kind of gives more clarity here.

It's quite interesting how the market's interpreting this earnings immediately, but the concentration risk is obviously still there. the the whether it's Microsoft, Google, Amazon, Meta, they keep paying I mean 44% of the what is it 200 billion no like 180 something billion they've done this year in revenue um was from the hyperscalers and so at the end of the day that concentration risk is there but they continue to spend for a reason they need the chips you're seeing the AWS GCP Azure growth rates so I don't know if that is the biggest bare case anymore and then you have the commitments to make the entire our supply chain makes sense.

keep paying I mean 44% of the what is it 200 billion no like 180 something billion they've done this year in revenue um was from the hyperscalers and so at the end of the day that concentration risk is there but they continue to spend for a reason they need the chips you're seeing the AWS GCP Azure growth rates so I don't know if that is the biggest bare case anymore and then you have the commitments to make the entire our supply chain makes sense.

Jensen's not going to be speaking probably for another 25 to 30 minutes. They're going to have to get through CFO commentary. Um, if you were an analyst on the call, anyone in the chat, and uh, my three fellow panelists here, let's start with you, Jose. What question would you ask Jensen?

Just give me visibility. That's it. That's all I care about because I think that ends a lot of the qu a lot of the other questions. yeah, and then on on on top of that, I don't even know how I would phrase this question, but based on all the commitments, it's surprising. I'm sure you guys have all seen the reports, the Vera Rubin Ultra is actually expected to have less HBM than the Vera Rubin. So, like, how does that tie in with some of their commitments? And um I I don't even know how I'd phrase that question, but like they are downsizing or or or lowering the specs on their future products, but also putting up unbelievable commitments to memory.

Is it a power play? Because if all of the customers/ competitors are also trying to catch up, they need memory. But if Nvidia picks everything up, they don't have the memory to move fast enough. And so it's like 4D chess from Jensen. Everything I do is a power play. Okay. That's just what Jensen does.

I would ask the most obvious question. What's going on with this open source thing? Right? Like I mean I would try to ask the question without being so direct, but I mean are you legit competing with OpenAI and Enthropic? Do you want more enterprises to adopt Neotron over the frontier models? And if they do that, is that bad for those companies? Will they adopt or is there going to be a sliding scale where some stuff is frontier, some stuff is open source? Um do is the open like look if if Nvidia, you know, is selling a decent chunk of open source tokens, is that another line of revenue for the business that is kind of uh latching on to the data center bit? I mean, I think the open source thing along with this commitment stuff that Jose talked about is a relatively big question.

I think the customer concentration risk, we've heard that question at least once every time over the past couple years. Um, I don't think that's going to be the biggest question anymore because these customers, Amazon, Google, Microsoft are showing massive cloud growth. So, they have to keep buying the GPUs. But that question might come out in the context of not customer concentration, but customer concentration and competition from those concentrated customers that are now seeing a lot of growth with tranium and with uh TPUs. Um, which again I don't think is the the most important question here, but I think we're going to probably hear some commentary around that. The stock is barely down though, so

It's just Nvidia right here. That's the one going down. I mean, granted, if Jensen says the right stuff and the street likes it, seeing this thing pop after hours, even though liquidity is not that strong,

we will be live after the call as Nvidia almost flips green right here to analyze everything that was said. And um there you go. It does flip green. We'll break down what happens. 210 on Mr. Invida. Now all the comments are going to be super happy whenever it opens green tomorrow.

She wants to know why Nvidia is down. Or actually, it's up now. I got everyone across the entire world watching this video 21040. All right, you know, this happened last time. It pumped just a little bit going into the call. So, now we just got to see if the call is good enough to keep it going.

driving a global infrastructure buildout supported by an expanding and diverse set of growth opportunities spanning hyperscalers, AI labs, AI natives, enterprises, and sovereign customers. We expect to grow revenue by approximately 70% in fiscal 2028.

This is a supply constrained outlook. Q2 data center revenue increased 18% quarter-over-arter to 89 billion with strong contributions from both sub segments hypers scale and ACE which includes our neocloud industrial and enterprise customers.

Hypers scale revenue of 49 billion grew 13% sequentially driven by sustained strength in Blackwell reinforcing that more compute drives more revenue as new GPU capacity comes online.

Our hypers scale customers delivered strong financial results in the quarter with accelerating revenue growth and expanding margins with cloud industry backlog now greater than two trillion.

Papex by the top five hyperscalers is expected to reach nearly 800 billion in 2026 and 1.3 trillion in 2027.

Today we are delighted to announce an expansion of our partnership with AWS. Building on its already vast installed base of NVIDIA compute. AWS is deploying an additional 2 million GPUs starting this quarter through the second quarter of fiscal 29. along with Vera CPUs, some integrated with Reuben, others standalone.

ACINE revenue of 40 billion increased 25% sequentially and 138% year-over-year. Growth was driven by NeoCloud capacity additions to meet the rising demand from enterprises, AI startups and sovereigns as well as hyperscalers purchasing capacity to supplement their own buildouts using NVIDIA DSX reference designs.

Our Neocloud partners are bringing capacity online faster and at lower token cost. They are expected to exit the year with 8 gigawatts in total installed capacity, up from approximately 3 gawatt at the end of 2025.

Incredibly, we are seeing demand acceleration even at our scale. Customers forecasts point to our growth doubling next year. However, as I mentioned earlier, we expect to grow approximately 70% as we are supply constrained.

NVIDIA compute is fully utilized across every cloud we serve. The economic value it generates for our hypers scale neocloud and AI lab partners keeps rising.

Besides building the best AI computing technologies and the most capable supply chain, NVIDIA has three unique capabilities that are engines powering our growth. First, Nvidia's architecture runs every model and we're growing share as closed and open model adoption grow.

Closed and open models alike, adoption is skyrocketing. NVIDIA runs the leading closed models, OpenAI, Anthropic, Grock, Meta, Gemini, and the leading open models, TML, Mistral, Quinn, Kimi, GLM, Deepseek, Miniaax, and Neatron.

We're great at small models and giant ones, large or video, auto, regressive, or diffusion, in the cloud or in the edge. NVIDIA is great at training, great at inference, great at agentic workloads. One platform fungeible for every model and workload, durable for the entire life cycle of AI. That combination of performance, fungeibility, and durability is what makes Nvidia the productive and financable compute infrastructure.

Our second unique capability is our full stack AI factory platform that is expanding our share of the data center town. Since Hopper, our revenue opportunity has grown from roughly 18 billion per gigawatt to 25 billion with Blackwell to 40 billion with Vera Rubin, which now spans Vera CPU, Ruben GPU, Envy Link, Infinibanner Ethernet, and Gro LPU announced earlier this week.

Our ability to extreme code design across GPU, CPU, NVLink scaleup networking, scale out networking systems, algorithms and software enables us to deliver X factor performance gain every generation.

Vera Rubin exemplifies this, delivering 30x higher throughput per megawatt and 35x lower token cost relative to Grace Blackwell Ultra.

We commenced production shipments of Vera Rubin earlier this month. Having already received purchase orders from every major hyperscaler AI cloud and system OEM, we expect Deer Rubin to mark the fastest product ramp in Nvidia's history.

Our networking business had another record quarter with revenue growing 18% on a sequential basis. Spectrum X Ethernet, which grew 2.6x 6x on a year-over-year basis is already helping us become the largest and fastest growing network company in the world.

Rising adoption of Agentic AI is driving an acceleration in demand for data center CPUs. Our Grace CPU introduced in 2021 has been a great success with revenue on a trailing 12-month basis exceeding 5 billion.

Today we are in full production of our next generation Vera CPU. As a standalone product, Vera expands our TAM even further. Vera completes a genic task 1.8x faster on the spec benchmark and provides five times the bandwidth per watt than any other data center CPU.

We expect Vera to be deployed by every major hyperscaler, Neocloud, AI lab, and system OEM with shipments already underway to our lead partners including OCI, SpaceX AI, and starting this quarter, AWS.

We continue to see demand for approximately 20 billion in total server CPUs. And based on our customer demand and improving supply outlook, our preliminary expectation is for CPU revenue to more than double in fiscal 28, positioning us as one of the world's leading server CPU suppliers.

Since the announcement of our Grock partnership last year, we've been working to unite Nvidia's high throughput and Grock's high interactivity architectures. At hot chips earlier this week, we announced that Gro 3 LPX, our first rack scale LPU system, is in full production and already setting records, demonstrating nearly 4x the number of tokens per second against the next best alternative on our artificial analysis benchmark.

We expect to ship Gro 3 LPX in volume later this quarter to early adopters. NAUAS will be the first. Today, we're not just selling the best chips. We're selling a full stack AI factory platform offering superior economics for customers and capturing a bigger share of the data center town.

Our third unique capability is the combination of our full stack AI factory and rich CUDA ecosystem allowing us to extend AI into markets a single chip alone can never reach. Beyond the hyperscalers lies a massive market anxious to adopt AI. Customers with no interest in designing their own custom silicon. NVIDIA's fully proven full stack platform is uniquely suited to help sovereigns, neoclouds, and enterprises build their AI infrastructure, bring it to full operation, continuously optimize it through CUDA software and connect it to offtake demand from our vast developer ecosystem.

Hyperscalers will remain a major growth driver but non-hyperscaler growth R AI and E segment spanning sovereign regional Neoclouds enterprise edge and air gap data centers will represent roughly half of our data center business.

Our AI native startup ecosystem developed and running primarily on the NVIDIA compute platform is scaling at a rapid pace. Global VC funding in AI roughly 70% of which is spent on compute exceeded 400 billion in the first half of 2026 surpassing the 265 billion raised in all of 2025.

Nearly 20 companies, including Cursor, owned by SpaceX, Figma, and Together AI, now exceed 1 billion in annualized run rate revenue, up from 13 companies in Q4 of last year, with vertical enterprise software logging the fastest growth in enterprise.

On a trailing 12-month basis, on-prem revenue in the automotive vertical reached eight billion, while financial services, manufacturing, and healthcare combined contributed 7 billion in revenue.

Hudson River Trading and Jane Street are leveraging Nvidia's powered AI factories to accelerate quantitative trading. Samsung Electronics is using Nvidia Kulitho to achieve up to 20x greater performance in computational lithography. In sovereign AI, our business primarily through the regional neoclouds grew 35% sequentially and more than tripled year-over-year in Q2.

A country or region can allocate land and power directly to a regional cloud partner in ways it never would to a foreign hyperscaler. We don't own a cloud ourselves. We are neutral partner to every sovereign and neo cloud. And because Nvidia compute is productive, fungeible, rentable and durable, regional cloud interest is surging around the world. We helped Quarweave, NAUA and NScale build entire infrastructure businesses and Neoclouds are emerging everywhere. Firebird in in Armenia, Casaba Technologies across Africa, GMI Cloud in Taiwan, Yoda and Nasa in India, Hermis in Australia, YTLAI cloud in Malaysia, pairing local land power and operating expertise with our platform. Last month, we announced a partnership with Noatra, Japan's national AI company, to build an NVIDIA DSX AI factory that will create open models to power AI agents, digital twins, robotics, and physical AI applications. South Korea's LG and Hyundai Motor Group are partnering with Nvidia to build and scale AI. And in Europe, a record 35 new NVIDIA powered AI supercomputers were unveiled to advance industry and scientific breakthroughs. NeoClouds are seeing strong demand pipelines for many diverse offtakers. Rather than allocating their entire capacity to a single long-term offtake guarantee that lenders typically require to finance a data center independently, we have introduced a revenue sharing structure. NVIDIA provides a take-pay commitment on a portion of the facility's capacity, a minimum revenue guarantee that gives lenders the confidence to underwrite the project, and in exchange, we share in a portion of the Neocloud's revenue earned above that floor.

Independent Capital still underwrites every deal on its own merits. We're not making loans. In this model, we get paid twice. once on the hardware sale and again through the share of rental revenue. A highly reoccurring stream layered on top of a one-time equipment purchase. Over time, this model can expand our addressable market and create reoccurring usage linked revenue stream alongside our core platform revenue with the potential to drive billions in revenue over the medium to long term. Together, Nvidia's three unique capabilities. A platform that runs every model, a full stack AI factory platform capturing more of the data center TAM, and a CUDA ecosystem that extends AI into markets no single chip could reach alone reinforce one another and are the engines of our growth. Let me update you on our progress with our Frontier AI labs. The Frontier AI labs have extraordinary demand for training and inference compute, but they are growing faster than what their balance sheets and credit profiles can support. They have rapidly growing customer demand, yet still lack the decadesl long infrastructure contracts and investment grade financing capacity needed to secure the AI factory infrastructure independently. In other words, their growth isn't limited by their technology or customer demand. It's limited by compute. For these companies, more compute means more more intelligence, more users, and more revenue. NVIDIA is needed to help power this flywheel. First, we've invested nearly 50 billion in the Frontier AI labs. This was a meaningful commitment, but it represented a small fraction of our expected free cash flow over the same period.

Further, to support the Frontier Labs infrastructure buildouts, we recently announced partnerships with six of the world's leading infrastructure capital providers, Apollo, Black Rockck, Blackstone, Brookfield, Goldman Sachs, and KKR to establish financing platforms that will raise over 500 billion of thirdparty capital. With these partnerships, building on our unique, funible, and durable computing platform, the AI labs will be able to build and assess AI infrastructure funded by long-term institutional capital at relatively attractive rates. Last week, we announced that we secured land power shell capacity through our partnership with SoftBank Energy to exclusively host NVIDIA compute at their Portsouth campus. The initial deployment expected to support 4.25 gawatts of AI factory capacity will be utilized by OpenAI. Each generation of NVIDIA AI factory systems deployed at Portspike could represent approximately 1.5 million NVIDIA GPUs and over 20 years the site could support multiple upgrade cycles. Here's the essential economic point. The LPS commitment secures a long lived AI factory site while the Nvidia compute within the data center can be upgraded repeatedly.

This project deepens our long-standing partnership with OpenAI. OpenAI has committed to substantially deployments of NVIDIA AI infrastructure through 2030. OpenAI's existing and planned commitments represent approximately 12 gawatt of Nvidia compute. for another Frontier AI lab. We will provide selective credit enhancement for nearly two gigawatts of compute. This complements the substantial NVIDIA compute capacity they've secured independently without Nvidia's credit support. We recognize the scale of this support and we know some will call this circular financing. We see it differently. We're going through a major computing platform shift, the creation of one of the most important technologies in human history, and these are once- in a generation companies. Their technology leadership is proven, and their customer traction and usage are skyrocketing. We expect them to become the largest technology companies in history. We believe these investments measured against the strength of their demand, the business they create for us, the ecosystem they build on Nvidia's platform, and the equity returns on our invested capital will be excellent, and our risk is limited. The NVIDIA compute platform is funible and durable and can be redeployed to support other customers. For context, we expect demand from the AI labs for which we expect to leverage our balance sheet to contribute toward roughly a quarter of our business next year. This remains compute we ship will be consumed by investment grade customers or those that are backed by one.

In Q2, we ship less than 1% of our total data center revenue in Hopper 200 products to customers based in China in accordance with the US government licenses. Current Hopper shipments are dilutive to corporate gross margins and given ongoing geopolitical in uncertainty, there is no China data center compute revenue in our forward outlook.

Moving to the rest of the P&L, GAP and non-GAAP gross margins were both 75% largely unchanged from last quarter due to a similar product mix. GAAP and non-GAAP operating expenses were up 10% and 11% sequentially primarily due to high compute infrastructure costs and compensation and benefits costs. Our non-GAAP effective tax rate of 16% increased from a year ago primarily due to higher revenue.

On our balance sheet, inventory increased to 32 billion as we prepared for the Vera Rubin launch. Days of sales outstanding increased to 60 days, reflecting extended payment terms for large purchases by certain investment grade customers to be shipped over multiple quarters.

In Q2, we returned a record 26 billion to shareholders, 20 billion through share repurchases, and six billion through our quarterly dividend of 25 cents per share. Relative to our plan to return 50% or more of free cash flow, we have returned 60% on a year-to-ate basis. And going forward, we intend to increase and return excess free cash flow net of strategic uses.

Let me turn to the outlook for the third quarter. Total revenue is expected to be 108 billion plus and minus 2%. We expect sequential growth to be driven primarily by ACINE with data center while growth in hyperscale is expected to reacelerate in Q4 and into fiscal year 28 as supply of Vera Rubin grows over time.

We see Vera Rubin accounting for about 20% of data center revenue in Q3.

Looking ahead, our preliminary expectation is for fiscal year 28 revenue to grow approximately 70% year-over-year. Although we will work to close the supply demand gap, we expect supply to remain a bottleneck at least through the end of fiscal year 28.

Many of you have expressed concerns regarding our gross margins as component costs have risen significantly. As you are already aware, we are experiencing extreme pricing conditions in memory. The magnitude of the price increase has exceeded our prior expectations and are headed even higher into next year. As a result, we are resetting expectations today. For Q3, we expect gap and non-GAAP gross margins to be 74% plus or minus 50 basis points. We expect margins to bottom in Q4 in the 71 to 72% range

We want to be direct about this rather than let it linger as an open question. Memory scarcity today is being driven in large part by the AI buildout itself. And unlike a component that simply raises our cost with no offset benefit, tighter memory supply is a symptom of the same demand surge that's driving our own growth. We have a longstanding deep relationships with all three major memory suppliers, and we're working closely with them to further increase the capacity our roadmap requires.

GAP and non-GAAP operating expenses are expected to be approximately 9.2 billion and 9.0 billion respectively for the full year. We now expect OPEX to grow in the low 50s driven by a broadening of our product portfolio and further increase in the usage of AI tools which is already and will continue to enhance engineering productivity

for full year fiscal year 27. We continue to expect GAP and non-GAAP tax raises to be between 16 and 18%. Excluding any discrete items and material changes to our tax environment.

Yeah, thanks Joe. Um, as as you probably are aware, AI has become useful and the the AI agents uh that are being adopted everywhere use an enormous amount of compute. Uh, first of all, the large language models are are larger than ever because they're smarter than ever and and these agents uh go through uh reasoning and planning, multi multiple turns of tool use. uh the amount of compute necessary for an agent versus a human using it is probably 15 to 100 times depending on on the type of problem you're trying to solve. And so the amount of compute necessary is just extraordinary. That's that's a factor that almost everybody sees. The part that people don't see about our growth because we're practically singular because the the nature of how we deliver products. I mean, we're the only company in the world that creates and builds offers an an entire AI factory platform, a full stack system. And you know, customers can still mix and match. However, uh most most companies just don't have the skills to do that or desire to do that. And so, there's an entire part of the market that we experience growth. uh their sovereign AI, the regional AIs, their neoclouds, their AI startups, their enterprises where where we're seeing um which is represents about half of our business and that's growing 100% a year. That part of the the world's computing is likely to to be larger over time uh than even what we're currently experiencing in the cloud. And so so I think the the demand that we see is driven by all of those factors. Uh it is also the case that that you can no longer procure uh technology per se and stand up these infrastructure. you've got to go secure the land power and shell which you know often times is a couple two three years out um all of the rest of the supply chain necessary to uh align the construction the power the cooling um you know all of the labor that's necessary and we're AI infrastructure is creating so many jobs all over the United States and all around the world uh it just takes a lot more planning and so we're involved in in securing infrastructure now uh further they're down the pipeline. You know, just as a long time ago, people ask me, why is it that we're working with memory suppliers uh when we're a chip company? And today, people understand it's it's really quite genius that we were working on our supply chain so far upstream. We work with power generator companies. We uh downstream we work with um uh work with land power and shell companies um all around the world. And that helps prepare all of this computing that's going to be built uh that will ultimately deploy for our ecosystem and our customers. And so we just have a lot greater visibility now upstream and downstream. Um it is the case that we've never forecasted or never guided to uh a year in advance and and um even though our demand is much greater than 70%. Um, our supply allows us to confidently deliver 70%. And we're going to continue to work with our supply chain to increase on on that. Uh, but what we wanted to do is to be consistent with everybody from our customers, our shareholders, our supply chain, everybody sees the same view. And the reason why that's important is because, you know, everybody's putting a lot of resources at play. And so we wanted to make sure that everybody has the same set of information. And um we've got a huge year coming up next year. Uh and and uh it's going to be pretty extraordinary.

Yeah, thanks CJ. um the AI life cycle is getting way more complex than it used to be and and it's playing um you know into Nvidia's architecture much much more greatly than it used to be and and so you could kind of see it as four phases you know there's the first phase which is preparing all of the the data that you need some of it is synthetic some of it is real some of it is human labor human labeled and generated pre-train the models and then there's post training the third phase and then there's the agentic inference and agentic inference is extremely complicated and so every one of those phases are complicated. The thing that's really great about the NVIDIA architecture and we created this uh with MVLink 72 it was a big surprise on the world when we first created the first the world's first rack scale architecture it was it was it was hardly easy um and it was very challenging building the first generation we're now in our third generation of MVLink 72 rack scale systems we had to reinvent the entire supply chain reinvent systems reinvent the technology redistribute our software refactor factor, our software, everything, every aspect of was hard. But what it allowed us to do was to create one fungeable system that allows us to transition from data creation, data preparation, the pre-training to post-raining to agentic inference. The the benefits to customers is incredible. And the reason for that is because you've just spent, you know, and we just mentioned um each gigawatt of technology and Nvidia's revenue exposure in the hopper time frame with hopper plus infiniband and now vera rubin and CPU and three types of different networking because it takes that many types of networking to address the entire world's data center not to mention the scalein security networking and the scale across multiacus networking. So you could you could argue five different types of networking systems and then of course Grock and all of that increased our revenue contribution or revenue opportunity per gigawatt to $40 billion. So each gigawatt of data center um increased from say $30 billion about five years ago to now $60 billion uh today. Of course, the productivity is tremendous. The the performance is incredible uh in comparison, but you're talking about a 60 billion dollar investment. And to the extent that you could you could use it across multiple phases of the AI life cycle, run every single type of model you can imagine running on it, whether it's diffusion or auto reggressive or state space or some hybrid version of that. every version of attention mechanism you can think of small or large models um the light the the the sa the the the investment that you make um will be pres preserved and useful and productive for a lot longer time and so so I think our our advantage in this new world is really quite extraordinary and it it could explain you know why it is that our growth is actually accelerating it was already large um but now it's accelerating

Uh let's see you as you asked about Grock. Super excited about uh Gro 3. Uh we achieved a uh record token interactiv interactive interactivity rate extremely low latency performance generation. The team is doing fantastically. uh we spent the last several months fusing uh the MVLink architecture uh which will be the the core and it'll be the core engine and then for for uh services that would like to have super high interactivity uh super high speed token generation done um you know the the throughput is going to be a lot lower uh the cost per token will be higher uh but you could you could associate it with you know high ASP uh services and so for those companies you could bolt on one of our Grock accelerators. I'm super excited about that. But the vast majority of the world's data centers will just be Vera Rubin MVLink72.

the unconstrained would be lot a lot higher. It's um we grew 100% year-over-year. this year. Um the unconstrained uh you know is is significant. And so we're just going to have to go work hard to get more capacity. Uh and and you know we have we have a large supply chain. We have a really gigantic supply chain and and so we have incredible partners um and we've secured a lot of supply but we just need a lot more. uh to break it down the the way to think about that is most people see just hyperscalers and that's half of the that's half of the picture the other half of the picture is what we call AC and that's all the enterprise the neoclouds the sovereign AIS you know that part of the world is invisible to everybody and the reason for that is because they they don't buy custom chips they don't buy chips one at a they really need an entire factory platform built for them. Um and and so that's that's a space that we add just a tremendous amount of value. Now of course back in this hypers scale space that's growing incredibly too right you know that they they now have backlogs of $2 trillion you know that uh when they stand up Nvidia compute when that happens their revenues go up their earnings contribution go up uh compute is profitable very profitable today and compute directly translates into increased revenues and so there's a there's a just a a race to want to bring more Nvidia compute online um both at the hyperscalers, but what you don't see is just really tremendous opportunities outside of the hyperscalers. Um but the the other the other part of it is and the reason why you know we mapped it out for you uh in the case of Hopper we were at about 18 billion per gigawatt. Uh for Grace Blackwell we're about 25 billion per gigawatt and for Vera Rubin it's about 40 billion per per gawatt. and and um the productivity is you know X factors uh increase in each generation and so customers want to race to the next generation as fast as they can. Meanwhile, because Nvidia's compute is so productive, they're the tokens they're generating, the GPU hours they're renting out is insanely profitable. As you know, their margins are fantastic. And so so all of that is just simultaneously happen happening. I think the big picture is is that we're going through this platform shift and it affects every computer company and every comp every industry in the world uses computers. So therefore every industry is affected, every company is affected and this new way of doing computing um is intelligent. It's not based on retrieval of files, but it's now generative generating intelligence and that requires compute, but the results you get is phenomenal. The results you get is tremendously better. And so, you know, you wor yet we're just seeing that across the world. Everybody wants to be part of the AI revolution. Everybody want everybody will have to be part of this computing shift and everybody has to build infrastructure.

happens their revenues go up their earnings contribution go up uh compute is profitable very profitable today and compute directly translates into increased revenues and so there's a there's a just a a race to want to bring more Nvidia compute online um both at the hyperscalers, but what you don't see is just really tremendous opportunities outside of the hyperscalers.

Um but the the other the other part of it is and the reason why you know we mapped it out for you uh in the case of Hopper we were at about 18 billion per gigawatt. Uh for Grace Blackwell we're about 25 billion per gigawatt and for Vera Rubin it's about 40 billion per per gawatt. and and um the productivity is you know X factors uh increase in each generation and so customers want to race to the next generation as fast as they can.

Meanwhile, because Nvidia's compute is so productive, they're the tokens they're generating, the GPU hours they're renting out is insanely profitable. As you know, their margins are fantastic.

And so so all of that is just simultaneously happen happening. I think the big picture is is that we're going through this platform shift and it affects every computer company and every comp every industry in the world uses computers. So therefore every industry is affected, every company is affected and this new way of doing computing um is intelligent. It's not based on retrieval of files, but it's now generative generating intelligence and that requires compute, but the results you get is phenomenal. The results you get is tremendously better. And so, you know, you wor yet we're just seeing that across the world. Everybody wants to be part of the AI revolution. Everybody want everybody will have to be part of this computing shift and everybody has to build infrastructure.

Well, we're building something very different, you know, whereas whereas uh many of these these XPUs are inference specific chips um for one cloud or one service, NVIDIA is a a platform an entire AI factory platform that spans the entire AI life cycle that you can use in any cloud. It's in every cloud. You can run anywhere. We'll help you set it up anywhere. And um I and so we built something very different. Um these all of the AI services at some point are going to want to go around the world and those data centers won't necessarily be just built by them. and and um and also you know I fully expect and so they're they're going to run and I think they're going to run on Nvidia u all around the world and and of course I think our technology I have 100% confidence that our technology will continue to be extraordinary for them and that the economics of using our technology you know whether it's from data processing to to uh training to post training to agentic processing um our technology is going to be extraordinary for them. They're going to use it and so I have every confident that they're going to be customers and partners of ours for a very long time.

Now, having said that, taking a step backwards, investing in these two companies or there are several AI labs that we've invested in, investing in these companies are once in a generation opportunity. I think the only regret that I have is that I didn't invest more and sooner and both of the two of the companies will likely go public soon and others will follow. Um, and these will be some of the most consequential technology companies in history. And so I'm delighted to be friend to be their friend. I'm delighted to partner with them. I'm delighted that they're building an ecosystem on top of the NVIDIA architecture. Uh, I'm delighted that that they're they're counting on us to scale up. And I I I have a 100% confidence um that you know through through quite a long period of time uh they're going to be utilizing Nvidia compute uh for a lot of their computing and so I feel great about it.

>> So let me add a little bit more regarding uh the commitments and uh the portion within those commitments which is our supply commitments. This is essential. This is essential for um the raising of Vera Rubin today as well as all next year. You can see that those commitments the biggest parts of them are in the first three years and we will use that to build the products that we need. This what also gives us the confidence in terms of our growth in revenue given how much we have already aligned in commitment in terms of our supply as well as capacity that we would need.

Um the world will need both closed models and open models and both closed models and open models are skyrocketing in use. um near most I would say nearly all open models run on Nvidia and the reason for that is because Nvidia's footprint around the world um is the highest and and um our our architecture is the most funible. It's everywhere. It's in PCs and edge devices like DGX Spark which is doing great all the way to robots and workstations and your on-prem data centers.

Um, open models are doing incredibly well. You know, closed models we we know are doing incredibly well. Uh, the Frontier Labs their scales their sales are skyrocketing. Their margins are fantastic. They're they're generating profitable tokens. They're only limited by the amount of compute. That is equally true for open models and our position in open models is very good you know because the CUDA ecosystem is literally everywhere. The open model open models are also foundational to just about every AI startup and every enterprise company around the world. It's vital to them. And the reason for that is because you should rent intelligence, strong intelligence, smart intelligence wherever you can, which is the reason why, you know, we rent it and I encourage my employees to use the cloud services as much as as they can. But every every major company and every surely every country and every startup needs to build their domain specific their proprietary proprietary AI their proprietary alpha and and the open models reaching frontier levels has made it possible has enabled them to all do that. One of the areas where frontier models is vital is cyber security. You see the number of cyber security companies that are enabled by frontier models so that they could have distributed massively distributed continuously running autonomous cyber security systems to defend. Um it those companies are emerging there's some amazing companies they couldn't do it without open models. And so open models is is both uh incredibly successful and has finally reached the frontier, but it's also vital to the American economy. It's vital to the world economy. It's vital to companies to build their own proprietary AI. You can't do it without one or the other. Both are going to be extraordinarily successful.

And and lastly, uh, as you know, uh, our market footprint of of all AI models, we're the only, I think we're the only platform, I'm fairly certain, we're the only platform that runs every Frontier model and whether it's closed or open. Um, most of them were built on Nvidia. And so they run they run great on Nvidia and and so we're delighted um by by any model succeeding so long as models succeed. I'm very happy and and both closed and open models are going to succeed and they're they're both simultaneously driving our sales.

>> I appreciate that. It's going to inflect further. Um today the vast majority of AI is prompted by people. I believe that this last month it has crossed. Um most AI are now agentic. But in the future in the future every company will have a whole bunch of agents. You know we have we have 40,000 employees roughly. In the future we'll have 400,000 agents, four million agents. And those agents are running continuously. They're running in the background. Um, if you have anybody who builds, if you know anybody who builds edge, you know, personal AI agents and they run it on their like a DGX Spark and you know, I I know a lot of people who who uh run it on DGX stations, this incredible workstation that that we've built and you can buy it from Dell and and they're they're they're incredible. And these AI agents running on a a DGX station runs 24/7 because you got stuff for that to do for it to do all the time. And so the when when the world goes to agentic fully agentic agentic systems, you're going to have agents running all the time, working with other agents running all the time. And those will be working in the background improving your company, you know, improving your lives. Um, in a lot of ways, we're we're kind of recursive at this point at this point. And you could argue it's it's coarse grained. Um, but every time you run through an agent, it it reflects on how it could do a better job next time. And it updates the skill file. And so the skills document, the markdown is updated at the end of every single one of them. And so next time you run it, it's going to get better. It's a it's a clo it's a, you know, kind of a loosely coarse grain self-improvement. And so you see that all all over all all always all ways. and and so in a lot of ways and for many for many tasks um we could say that we've already achieved AGI. I think all of that those milestones and all those you know they're kind of senseless at this point. I think the most important thing that that that matters for the industry is that one AI is now doing productive and useful work. Two, AI is generating profitable tokens. And three, if we had more compute, we could generate more profitable tokens, which results in more profit for all of the services. This is the exact phase where we're at, which is the reason why everybody's leaning in.

>> There's fun something funny I could say but I'm going to just not um the the uh you know last year one of the funnest things to do is just to go figure out where I go for dinner and and who I have dinner with and um their their stock price doubles the next day. Um I I I think the the answer is is um our entire supply chain is challenged and it's everybody everybody is really running flat out. Um and more capacity is coming online all the time which is one of the advantages what's going to happen this year. It's not going to come online in just you know an instance in time but it's going to come online every every day. Yields are going to get improved. Uh we're going to be doing yield improvement. uh we're going to work hard on on um working with every one of our suppliers, you know, and so we're just we have a it's not even next year yet and so we've got lots and lots of time work hard every day and so um at this moment uh we have supply for 70%. We have we have more supply than 70% but about 70%. our demand is much higher than that and we've got to go work hard or you know we're we're going to be uh disappointing customers and and um uh we we like not to disappoint our customers and we like to work hard for them and so I've got I'm going to need the help of the entire supply chain to help me out here. But they all know that um uh what what I'm telling you about about our needs for next year uh is exactly consistent with what I've told them. Um everybody's on the exact same song sheet song sheet and I'm trying to be as as transparent as we can because we're talking about big numbers.

>> Your final question comes from the line of Aaron Rakers with Wales Fargo. Your line is open. Yeah, thanks uh for taking the question. I want to go back to the the gigawatts the 25 to 40 uh and maybe try and understand like you know I think Jensen you said at some recent conferences that you know that's going to further scale so you know as we think about the path even beyond ver rubin we think about ver rubin ultra and so on so forth like should we really conceptualize like 40 billion goes to 60 billion 80 billion and then you know I guess you know underneath of that question is how do we kind of think about your ability the the scale, the capacity deployments, you know, is it is it a linear function or is there something that, you know, kind of unlocks your ability to supply more demand as we look through fiscal 27?

>> Great question. >> 28, sorry. >> Yeah, great question. Great question. Um, and very simple. Is our goal to put as much compute on a plot of land? Is our goal to put more compute into one gigawatt or less? And so obviously we would like the you know the speed of light answer the perfect answer is actually infinity per gigawatt and and so if we could literally get a trillion dollars of compute into one gigawatt and one piece of land power shell it would be a fantastic outcome and so the answer is directionally in that direction. Uh we started we started in the world of general purpose computing during Moors law. Uh we were probably you know pick your favorite number but I'm going to go with something like five billion3 billion per gigawatt of compute uh with with general purpose computing and and then eventually with hopper it was 18. Uh now Grace Blackwell is 25. Next Vera Rubin is 40 and after that it's going to be higher and that's excellent. That's fantastic for the industry. It's fantastic for customers. So long as the productivity of it continues to grow, the durability and the fungeability continues to grow. Then people are happy to invest uh in in assets that generates revenues, generates profits and helps them recoup their return so incredibly fast. I mean, I heard the other day that return on investment capital is now less than a year,

Um, and very simple. Is our goal to put as much compute on a plot of land? Is our goal to put more compute into one gigawatt or less? And so obviously we would like the you know the speed of light answer the perfect answer is actually infinity per gigawatt and and so if we could literally get a trillion dollars of compute into one gigawatt and one piece of land power shell it would be a fantastic outcome and so the answer is directionally in that direction.

Uh we started we started in the world of general purpose computing during Moors law. Uh we were probably you know pick your favorite number but I'm going to go with something like five billion3 billion per gigawatt of compute uh with with general purpose computing and and then eventually with hopper it was 18. Uh now Grace Blackwell is 25. Next Vera Rubin is 40 and after that it's going to be higher and that's excellent. That's fantastic for the industry. It's fantastic for customers. So long as the productivity of it continues to grow, the durability and the fungeability continues to grow. Then people are happy to invest uh in in assets that generates revenues, generates profits and helps them recoup their return so incredibly fast.

I mean, I heard the other day that return on investment capital is now less than a year, and we're talking about $50 billion data centers. And so, that tells you something about about the the uh uh the productivity of MVS technology and the rentability of it.

Thank you. Uh before we close, please note that Jensen will be participating in a keynote fireside chat at the Goldman Sachs Communicopia and Technology Conference in San Francisco on September 10th. He'll also be giving a keynote at GTC Berlin on October 21st.

that was the call. Obviously, when we got the numbers, the probably craziest numbers we've ever seen from a company. Uh stock fell 2%, it has reversed. We got above 220 for a little bit, but we've been getting rejected at 219, up 4% after hours.

I thought it was one of the best calls and videos ever had. I thought the questions were answered uh in a way that really I think instilled confidence in the investor base.

Um A I agree. think I I was mentioning Chad that this is probably my favorite earnings call since kind of this whole AI cycle started and kind of when we started uh the call and kind of saw the commitment numbers. The thing I wanted to hear more about was um visibility and Jensen I I think that was kind of the first line Colette started with. Uh she mentioned that for fiscal 2028 which is calendar year 2027 they expect at least 70% year-over-year growth. Now, Wall Street expect is expecting about 400 billion this this year in revenue. And if you put that 70% year-over-year growth baseline, you're closer to $680 billion in revenue. Analysts were expecting about 580 billion consent uh on average. So, you're you're talking about an extra 100 billion. Um so, I I think the visibility question was answered there. So, to me that was the biggest talking point.

And after that I was just smiling and happy and it was the greatest earnings call ever because we got that visibility. Uh but the second thing that really grabbed my attention was um the 70% growth was mainly because of constraints in the supply chain. So then if you trickle down, you guys know I'm extremely bullish in the Neocloud market. If you are constrained in the supply chain, that means there's going to continue to be a supply demand imbalance in GPUs. And if there's the continued demand uh imbalance in GPUs, uh that means that these Neocloud players are going to continue to have strong pricing power on their end.

Uh so if there's two points obviously we can come back to to me later but if two points is the visibility question was answered right off the bat and the second thing is I expect the revenue per megawatt to continue to increase if we continue to see this uh supply constraints there.

Excellent. Excellent. I mean even even when they talked about the memory right the the fact that memory is the the the reason why margins are going to come down a little bit stock didn't really move that much they explained everything quite well they even mentioned circular financing they explained that quite well

and so yeah it was yeah it was an an excellent quarter great call stock is up so everybody's happy but the yeah they are doing exactly what they should be doing they also explained it quite well open source open way frontier model how they benefit from whoever wins Nvidia wins so it was a good one

Yeah, I mean they they covered a you know the main parts. Obviously Jose highlighting the guidance is by far the biggest point and that's why the uh stock completely turned around.

Um I think that there could be for real real possibility for Q4 of calendar year 2027 um of them potentially putting up a quarter of $200 billion because this is their own guidance. every single quarter they've had uh higher beats, quite large beats versus their own expectation, beating by four, five, 6% a quarter. Um so if you couple that up, this could be, you know, a fiscal year 2028 could be somewhere closer to about 700, maybe 750 if they end up putting up really really good numbers.

So um they said they're completely con supply constrained right now still. uh they said to to Stacy Raskin if they were unconstrained the numbers would be much higher. So as supply comes on hopefully um they can find out new solutions new breakthroughs to end up uh shipping more than they would have in the past.

One of the things that Jensen highlighted at the end there which I don't think people give nearly enough credit to is the payback period on these gigawatt data centers can be as low as under a year. And now the companies that have been able to put up those types of numbers are really signing deals for extremely short-term commitments.

Now, um I think a part that people should be highlighting here too is if Nvidia is going to be putting up these numbers, who else is here to benefit? The the Neoclouds that Well, Jose for sure. Jose for sure. But you know, you don't think Dell or SMCI or um a lot of the memory names that are continuously being a lot of the constraint here are not going to benefit. That's why we're seeing a lot of green candles across the board is the commitments that Nvidia just talked about is to the benefit of SKHEX to Micron to to many of these names.

So, um, while Nvidia is by far my biggest position, it's not the only way that I'm playing this boom, but there was not a single piece of commentary there from Jensen that spelled that this is even slowing down. So, look out tomorrow for a flurry of midterm upgrades for Nvidia's revenue in EPS. That's going to be wild. I if you thought some of the reports of them putting out $350, $300 price targets was exciting, wait till, you know, you get the revisions from 40% growth up to 70% growth. That I think is going to be quite exciting. That's why it's a disco in this room now. And um yeah, couldn't be happier.

This is a $300 stock trading at $200. I was just about to say I if we do not have any major macro concerns towards the end of the year. Uh I've been in the camp that it's hard for me to see the stock hitting 250 by end of year simply based on the nature of uh the size of the company with this type of growth and not just this type of growth but the guidance for how it's going to continue.

Notice when the stock took off and one of my friends text me that he wasn't watching the call what you know what what did it take to get the stock to moon. When Jensen said I think this was the fourth question he like kind of laughed about it. He was like, I think we will have demand grow more than 100%. He said that the exact quote was, "I think demand is just inflecting." Uh, and he sees visibility into more than 100% growth, which is kind of the the answer to that question of how would you grow if you weren't supply constraint. He was like, "A lot." He's like, "Right now, you know, we can only supply 70% of it. That extra 30% would be potentially hundreds of billions of dollars of revenue as Tanner is mentioning on a yearly basis."

They also gave annual revenue guidance for the first time this quarter. I mean, every level of transparency you wanted is there. I don't know how the street holds it under 250 going into the end of the year unless Ron Moore macro bond yields etc uh holds it down.

I mean because when you really think about what this trajectory looks like this is not a peak for earnings growth or revenue growth. This is almost just like the beginning. Now they're not going to grow triple digits for the next five years but it looks like you've got another year of this like 90 to 100% growth or at least 80 to 100% and then it starts tapering off by then. I mean, Nvidia would be trading at 12 times if it is it's flat and it's still growing at that level. So, yeah, I think this was one of the best calls we've ever heard from Nvidia. And if the stock sells off on this, I think there's going to be a lot of buyers that get ready to come in.

Jose, what did you think about this quote when Jensen was talking about the supply chain and uh how much more they have to invest in their ecosystem? The the this quote of going to dinner. Yeah. I just thought it was funny. I mean, I it's um but they continue to say it's they they go dinner with everybody. I mean they have such a there's so many players in the supply chain industry in the supply chain line for them that um they're going to continue to do these big investments. Um so I don't think I have much here to say I'm just yabbling here.

It's a contrast of the steak dinner from uh Karp. There was something I thought was pretty interesting. Um at the beginning I think Jensen did talk about how capex for the five hyperscalers is expected to be 800 billion this year in 2026 but for next year is expected to be $1.3 trillion.

Jensen is already kind of telling us what the capex of these companies are going to go and I do think Amazon is going to be a big player. I mean we saw that announcement. There's also another big announcement that Amazon and Nvidia announced. that the Amazon is working with Nvidia on Envy Link Fusion to kind of connect their 86 with Nvidia solutions, but it seems like Nvidia is also kind of leasing out or licensing out or renting out however their memory storage solution and Amazon is actually going to be a big port uh a big customer of that. So, I'm going to do a little bit more learning on that, but I thought that was interesting from the Amazon side.

Yeah. One one thing I'll highlight about the um capex numbers, there were some Goldman Sachs reports that said it would be 1.4 trillion next year and I think many many people brush that off and maybe they were slightly high. Um but the visibility that Nvidia has way more than a single analyst that's talking to, you know, their connects, they're literally putting in orders with Nvidia. So I would put on high like high regard that that 1.3 trillion is really what's going to come in. And and by the way, it could actually be much higher than that distributed across TPU, Tranium, Cerebrus, like they're not the only player. That's the visibility that Nvidia sees. So it there's a wild amount of spend that's expected to come in.

the visibility that Nvidia has way more than a single analyst that's talking to, you know, their connects, they're literally putting in orders with Nvidia. That's the visibility that Nvidia sees. So it there's a wild amount of spend that's expected to come in.

And I I was going to say I wonder if that 1.3 trillion is taking into account only a 70% growth due to supply chain. Like I what happens if that supply chain if Jensen goes oh I actually have extra supply that we ended up building something a lot faster is that capex actually closer to that 1.4 trillion.

Yeah I I think that's a great point Tanner on on the visibility. I mean we know that Google said 190, Amazon said 220. Meta is probably gonna be that tuna range. So will Microsoft, but it's probably going to be a bit more just like it was this year. Uh because, you know, Jensen's probably collecting the orders already and so are all the other stocks in the supply chain.

I also thought addressing the circular financing stuff head-on was really important. Colette, like I think this was in the first five minutes, she was like, "Some of you may call this circular financing. We see it differently." Now, you will have people that say you can call it whatever the hell you want to call it. It's still circular. But the fact that they are addressing this stuff headon and saying look this is what we're doing. She made it very explicit that these are not loans that Nvidia is giving. It's protecting their balance sheet but this is in this is important to proliferate the advance of AI.

She made it very explicit that these are not loans that Nvidia is giving. It's protecting their balance sheet but this is in this is important to proliferate the advance of AI.

Jensen gave you the same answer he's been given for a couple months now which is that you need both open source and Frontier Labs.

At the end of the day, it's important that those companies win uh for GPU demand to continue to be strong, at least in the short term for Nvidia. It's like the stock, you know, is lower than where it was last Tuesday.

I think if the NASDAQ was up a lot more after hours, Nvidia would be, you know, crushing 225 right here.

now Nvidia did say like because of the memory shortage that you know that's what's causing the pricing inflation but at the end of the day that's not bearish the memory names. The Neoclouds are ripping as Nvidia explicitly mentioned that the Neoclouds will be a big beneficiary of uh this new product that allows them to kind of split the revenue with them.

What's the chances that we get to 250 and I win this bet with Steve this quarter? Look, I think off the back of those earnings, I think 250 gets hit soon.

cuz I know this one was even more exceptional, but there is something about this name, whether it's 8% of the S&P 500 or 5 trillion that the street is not letting it m maybe you're maybe this is the one that breaks it.

But I feel like the reason why I say that really the I think the real difference is not just another quarter worth of earnings, but it's because of that additional next year guidance that really like the the upgrades I think are going to be really significant this time around.

what would really get us there more than Nvidia earnings is Trump announcing a ceasefire, like a legit one or a deal, full-on deal. it does feel like this company, I think Jensen will be on in about two minutes, is getting what it deserves here, at least to an extent after hours.

They didn't mention SpaceX. I was kind of surprised he didn't give a nod to Elon on anything. I thought he would given how much Elon is exclusively working with him.

Frankly, blowout you expect to grow about 70% in fiscal 2028. the street was looking for roughly 45%. What gives you that confidence? And how the heck could you do it?

Well, it's not easy, but uh let's say first of all, demand is super strong and incredibly it's accelerating. You know, obviously we're already a very large company, but to be able to grow continuously and now to accelerate our growth is pretty extraordinary.

Number one is that AI is now useful. It's doing productive work and the tokens that are being generated by these AI labs are now profitable. People are paying good money to use it because it's really productive. It's helping them, you know, do things and be more productive and and so the amount of tokens they want to generate is increasing because of use. The amount of tokens they're generating is increasing because the AI models are more complicated. And then now because it's profitable, they have so many more customers want to use it. the thing that's holding them back is compute. And so that's literally one pillar of what's happening in our business.

And then after that, uh we're seeing AI being adopted all over the world. Every single country wants to get involved. Every country needs to every company wants to get involved. Hyperscalers and the other half of our business is everything else. It's neoclouds and sovereign clouds and enterprise companies. And all of this is all growing at the same time.

And this is all happening of course at a time when we're rolling out Vera Rubin, our next generation product. It's going to be the fastest ramping product in our history. It's really exciting.

One of the things that's really different about NVIDIA is that we are funible, meaning that you can use NVIDIA across the entire life cycle of AI from data processing, pre-training, post-training, all the way to deploying the AI agents. That's number one. Number two, we literally run every single model. We're the only company that I know that runs every single frontier model, every closed model, every open model. It's built on Nvidia. It runs on Nvidia. And because of that, we're fungeable. You could use this across all these different use cases. And because we have so many developers, our ecosystem is so large and we're used all over the world. The asset is durable. It's long life. It's fungeable. And therefore, you know, it's rentable. People are renting it for great money these days. And so that's that's the reason why it's an investable asset.

It is true. In fact, uh this it's some of the CSPs have already been talking about that that that the the infrastructure that they paid for um paying Nvidia's uh prices that they built is now worth more than when they first paid for it.

Um and and the reason for that of course is that Nvidia's architecture is softwaredefined. Because of CUDA, our software is getting better all the time. We ship the hardware in the beginning. We we sell you the hardware, but we continuously improve the algorithms underneath so that the machinery gets better and better and better over time.

So over over the history, I think um Corey Weave, Mike over at Corweave recently said that A100 is going to be sold for nine years of its life. And the thing that's really incredible is during that nine years time, even though the hardware didn't change, we improved and introduced all kinds of new algorithms and software that we continuously enhance that infrastructure completely for free over over its entire life. And so that's the reason why the hardware what the infrastructure just seems to get better and better and better over time, just like just like fine wine.

Yeah, I think they're missing a very big point. First of all, um, this is the first generation of startups that needed tens of billions of dollars to get funded. When was the last time anybody heard of a startup that needed billions of dollars to get off the ground and needed tens of billions of dollars to become profitable? That just never happened. But that's really the nature of AI. The cost of building AI, the cost of deploying AI, just it's very capital intensive. And there are several companies, these Frontier AI labs that are once in a generation companies and we want to be investors in them. We want to support them. We want to be a partner to them. We would love for them to build their ecosystem on top of us, of course, and then scale up their business along with us. And so the opportunity to invest in them in the beginning was a great opportunity. Frankly, I just wish I invested more.

And now that they're scaling up, the tokens they're generating are profitable, their services are profitable, the number of customers is growing incredibly. The only thing holding them back out now is compute. But even during this time, they're not investment grade. They just don't they don't have the track record, the the capital track record, the financial track record to be able to capture or secure uh capital at a low cost. And this is where Nvidia could be helpful. we could be supportive of them to help them grow at a time when everything they make, their tokens generated are so profitable. This is really the time when we can help them with their flywheel.

And so, as far as I'm concerned, this is one of the best investments ever. It's a once in a generation, once in a technology generation opportunity. And then this lastly, the risk for us because of the nature of Nvidia's architecture because it's fungeable and it's durable. the risk to us is still in the worst case scenario, we'll just use it for another customer in another use case.

And so, um, I think the backs stop is is, um, uh, we're thoughtful about it. We're disciplined about it and we're supporting extraordinary companies as they're becoming profitable and as we know, uh, you know, there hopefully they'll be soon there'll be there'll be public companies and people will all get the opportunity to see, you know, what we see and why we're so excited about them. And but anyways, this this is a really great investment and I think the risk is is low.

You know, I'm okay with it, Jim. There's so many XPUs that are being announced and as we know, it's not easy doing what we do. We've been doing this for 33 years and so lots of projects get started, a lots of projects get gets cancelled. Um, we're we're here we're here to support our partners and and um we're going to build the world's best technology. I have every confidence in that. Uh, we're going to be the most productive infrastructure that they have. I have every confidence in that. Um, we have the supply chain and the technology scale to be their largest supplier. I have every confidence in that.

And so, you know, I I don't I don't have to take anything personally because I've got so much confidence in what we we're able to deliver. And look at look at all of the XPU announcements and all the startups that have been announced. And yet today, Nvidia is increasing our market share of the AI market. We're our growth is accelerating. Our technology leadership is extending. And so I'm very comfortable with all the competition.

They're going to buy 2 million GPUs. They're going to buy millions of CPUs and they're going to uh build their robotics fleet on Nvidia's physical AI systems. They're going to host Nvidia's Neotron open models on Bedrock and Sage Maker. Yeah. So, this is going to be a great partnership.

Um, you know, I I think it's just really important that people realize that Nvidia is the only computing platform that is a full stack AI factory. We're the only computing platform that is in every single cloud on prem at the edge and sovereign sovereign AI clouds and neo clouds. We're the only platform that's in that way that's everywhere. And because of that, we attract all of the world's AI startups.

In this last six months, $400 billion dollars of venture capital went into startups because of AI. This is a once in a generation technology platform shift. Everybody wants to be part of it. $400 billion dollars of investment into AI AI startups and all of them are probably using Nvidia. And so if any of the clouds would like to attract any of those AI startups, having Nvidia in their cloud is going to make make that their clouds much more attractive.

And so I and however you think about it, whether we have the best technology, the most productive technology, the most rentable technology, you know, in every single case, Nvidia is a great partner of yours.

Yours is selling arguably at 16 times 28 earnings. That seems ridiculous to me. Is it time to make that buyback much bigger than the roughly $98.5 billion left to buy? Why not double that? Why not do what Luca Mistri did? He's such a great CFO. You have such a great CFO, Colette Crest. Huddle with her. you know, just huddle with her and come up with a number more befitting of how of how cheap your stock is. Best investment of all these things you're talking about.

Why not do what Luca Mistri did? He's such a great CFO. You have such a great CFO, Colette Crest. Huddle with her. you know, just huddle with her and come up with a number more befitting of how of how cheap your stock is. Best investment of all these things you're talking about.

We we are going to I appreciate that. Um we are going to uh uh return cash to our shareholders. We're going to buy back stock um and buy back stock uh all the excess cash net of what we need for strategy and operations. And we are going to generate a lot more free cash flow next year than this year.

Furthermore, furthermore, um, this year we have several very large investments and those large investments in the AI labs are not going to repeat. You know, the opportunity for me to invest in these frontier AI labs is not going to happen again. And the reason for that is because I expect them to go public. And so our the money we've invested is going to get is going to generate tremendous returns. Um, but I doubt that I'll have the opportunity to invest more. And so next year we're going to have tremendous amount of free cash flow. Uh we this last quarter bought back 60%. And we're going to increase that uh going forward.

So so I I think uh I think we're we're we feel the same way. Um buying back Nvidia stock is a tremendous opportunity,

cheapest asset out there.

I mean, I thought the answers to every single one of those questions was in line with the earnings call. Just really, really good communication from Jensen here. Was the for for real the question around um water use and and power and the more environmental questions. How much is that an Nvidia question versus like a verdive question like you you know what I mean?

Obviously the GPU and the actual chip side does play a part but they're not playing the the major part here. I just think maybe, you know, Jim obviously wanted to ask that question that might be on people's mind.

Yeah, he's probably using because if Jensen speaks, it's it's Jensen. it's that's like if Jensen says something, then I assume the whole industry will just say, "Okay, you know, yeah, you're right. We should do it." Which I think they all do anyways. like they it's in their benefit to make sure that that all of this noise goes away which will it go away will it not I don't know in the US you guys have a midterm election later this year in two months or so expect the noise to increase significantly

I think I think the opening eye question was also really good I mean essentially Jensen basically said everyone's trying to do custom AS6 uh they're not going to compete at the size and scale that we are due to our supply chain, CUDA library, all this stuff. I don't think he's saying they're not going to have usable chips, but it's not really going to impact Nvidia's growth trajectory.

Yeah. He said, uh, you know, some of them show up with good results, some of them get cancelled. He's like, we don't even know if this is going to come to production. Whenever a company has this high of a margin of safety, I mean, just the the ease of holding this stock, I'm sure this goes for all of you guys, but this is not really the speculative position that causes you to not sleep well at night. I mean, whenever you start to put this up against other names that people get very excited about, and I'm not I'm not here to bash this name.

Nvidia just topped that doing $96 billion worth of pro highly profitable revenue. So, um I mean the size and scale versus other companies at much much higher valuations in terms of how much it is versus their revenue and versus their profit, it's just wild. This company is very cheap. Like uh Jim Kramer said,

Nvidia up 4.3%. the entire sector benefiting from Nvidia earnings. And seems like Jensen executed communicationwise and with the numbers that were just some of the, you know, most monumental numbers we've ever seen. And tomorrow in the market open, we're going to see how much the street actually respects these numbers.

All right, thank you all for joining another Nvidia quarter in the books. Deeply, deeply appreciate it. means the world. Thank you everyone for all the super chats on all the channels. And yeah, we'll keep covering the markets. We'll keep covering Nvidia. But it means a lot for all of us to come together, you know, thousands of people across the world uh tuning in to our analysis and coverage. It really is special and you know we deeply appreciate the opportunity to do this every every quarter for Nvidia.