저자: 미상 · 외국인
요약 보기
GTC 2026 개막 전 시장 점검 방송. 엔비디아 2.5% 상승, Nebius가 메타와 270억 달러 규모 계약으로 15% 급등. 마이크론은 공급 부족으로 메모리 가격 2~3배 상승 가능성을 언급하며 GTC·실적 전후 단기 트레이드를 밝힘. 이더리움 2300달러대 회복, BMNR 등 암호화폐 관련주 언급. 유가는 94달러대로 하락, 나스닥100 1.3%·S&P500 1.1% 상승.
다듬은 글
Ladies and gentlemen, ladies and gentlemen, welcome to the midday market monitor. In 1 hour, we are going to have Jensen Huang come on stage to to talk at the GPU Technology Conference. This should be very very exciting. They're expecting to show off their latest platform that is is expected to have products that were designed by Groq in their new what I'm expecting their Vera Rubin Ultra platform.
Very very exciting, but until then let's go ahead and watch the market a bit. Make sure that we're all caught up on all of the news. Nebius had major news alongside Meta, Micron, Coreweave, a bunch of the Neo clouds that are actually premium sponsors, elite sponsors of GTC themselves. So, this should be very exciting, but let's go have a look at the market.
Overall, Nvidia is up 2.5% some pretty nice candles actually on the day. So, I'm going to make sure that I follow that closely. Make sure that there's no news that just came out, maybe some leaks or something like this on what is to be talked about. Trump just said on the Iran war that they are in talks with our people and that we don't need oil.
We have all the oil we need. That was another that he just said, but they want to make a deal. Mhm. Okay. That's what he just recently said, but let's go ahead and take a look at some other names. Nebius, if you guys didn't see this morning, up now wow, 15.6%. Nebius ended up signing a two if the full contract of the deal ends up going through, a 27 billion dollar deal with Meta.
Meta originally came to Nebius with a 3 billion dollar contract and they wanted to do more, but Nebius said that they didn't have the capacity to take them on as a customer. Okay? Then they ended up getting that customer. Um obviously you know uh showing that they have good products and that they're building up data centers at the at you know, according to the timeline that they originally agreed upon and that was delivered on time.
So they came back and nine X'd their deal. Well, I guess total I guess a 10 X, but they turned a 3 billion dollar contract into a 30 billion dollar contract total or 27 billion dollars of new additional compute needs. Um this is very interesting and ended up bringing up the stock up 15.3% uh good timing for me personally, but also uh good timing for GTC right before Jensen sat up on stage one of their elite sponsors ended up uh showing this off and you can actually see this.
I thought that this was a pretty interesting part that uh Evan from Stock Market News ended up showing off. Maybe I can actually open this image, but this is who ended up showing this off the the the sponsors of this thing. Let's make it a little bit bigger. But you can see the elite sponsors AWS, Coreweave, Dell Technologies, Google Cloud, H uh HPE, Microsoft Azure, Nebius and Oracle.
Talk about the companies that have well, one that that the closest relationships with uh Nvidia, but also have extremely high compute needs. You have the hyperscalers Microsoft uh Azure or sorry, Microsoft Azure, Google Cloud and AWS by far the biggest players. The the leading neo cloud names by far the biggest Coreweave.
Nebius is there as well. Oracle obviously having probably maybe you can name them as being the fourth largest cloud behind the hyperscalers. They're growing in insanely quickly and then underneath you have a lot of very big names as well and there's other sponsors even under diamond tier which is like platinum and these sort of things.
That's where you'll find Iren. So I I Ren was not up there with the N scale and together AI and Vertiv and these sort of names that you'd also see in the data center space but overall pretty good list of names that are working with Nvidia. Then our king Nebius. Yeah, I know. It's it was a good timing on um on that report overall.
Um Sponsors are those who pay, no? Yeah, of course. Of course, but they want to have very strong relationships with Nebius. Like they understand that uh you know, they they want to be in the good graces of of Jensen Huang. He is a kingmaker. He really is. And we're going to see that on this keynote.
Like like potentially the new softwares that they end up talking about, they could bring up names like Microsoft setting up the first Vera Rubin platform or Nebius on the recent deal that they just had or or Coreweave. They're very very close with Coreweave. So we'll see. Okay. Uh how could it be that Hood is not up more when ETH and Bitcoin are pumping?
I think there's general skepticism in the name right now. Like like whenever we're in extreme fear, people are just getting concerned that equity volumes were down last month, that um you know, options contracts were down. Crypto's not the only source of business that they have. However, whenever we look at Ethereum, yes, we're into the 2300s.
I just posted a video on uh R this morning because of this recent change in uh ETH prices is allowing BM and R to essentially skyrocket. But, that's not the only thing that they've ended up doing recently. BM and R just recently did a uh uh an investment in OpenAI, essentially indirectly holding them through Orbs, as well as also uh in additional investment in Mr.
Beast, also through Orbs. So, Orbs is a company that is is publicly traded. You can follow them here, ETH Co Holdings. And um ETH Co Holdings, which is up 39% today. Wow. I didn't actually know that. That's crazy. So, they're up 39%. That alone ends up bringing up the the total price for like like the total holdings for um BM and R, right?
Cuz they just additionally put in in another 80 million dollars into that company. And then on top of that, whenever you get OpenAI or something before IPO, the price of that company could be in the trillions after they IPO and we end up seeing another raise, maybe this year, maybe 2 years from now.
Um So, it it it it was a decent find. But, but Orbs essentially follows Worldcoin. Worldcoin was created by Sam Altman. Sam Altman also um helped co-found OpenAI. So, that's how they got that um that's how they got that uh entry into that. Um let's check out the Mag 7s though. I see you guys talking about Tesla here.
Let's have a look. Um Currently, the the biggest name right now in I mean of course, is Nvidia up 2.5% Meta up an additional 2%. What's actually funny is they did this deal with Nebius and ended up popping on that news. The idea that they're working with Nebius and spending 27 billion dollars over 5 years is actually brought up the stock.
That's not usually uh what has happened. When's the keynote? The scrolling bar at the bottom, Nvidia GTC starts at 2:00 p.m. Eastern Standard Time. And you know what's so funny, guys? Uh blonde girl here says Micron. Guys, I bought Micron. I bought Micron this morning. Um we're getting ready, ladies and gentlemen.
I think not only GTC I mean there's a possibility that GTC could bring Micron down if they end up announcing some sort of uh change to the way that they need DRAM or something like this. There there there could obviously be some some worries with Micron, but I think Wednesday is going to end up showing off some pretty unbelievable guidance just because you're seeing the amount of acquisitions that are coming out from some of these bigger you know, memory names, but like like for example, Micron just acquired PMC uh Tonglo P5 fab in in Taiwan, 300,000 square feet of uh clean room space.
And then going forward, that's going to be an an additional 270,000 square feet of clean room space. According to Dylan Patel, that's the largest supply constraint right now. Um so so they're setting up these fabs, but the the retrofit's going to start in March 2026. They're not expected to see shipment until fiscal 2028.
This is a problem. There there's going to be a massive constraint here and the price of memory could two three x from here in that meantime. Okay? After two three x'ing. So, the the massive skyrocket in memory prices and the still what I think is uh a very cheap price for for Micron which honestly, I saw it much earlier than this and didn't pull the trigger on it and it was a massive mistake, but I'm not going to miss it this time.
I'm just going to have a trade, but this is for a trade on GTC and a trade on earnings looking forward to some pretty explosive guidance. That's what the trade is for. I'm not going to hold it much past that unless I think that it's going to be a six-month or nine-month build-out, but that's pretty much it.
But I honestly think that the market's smart enough to realize that it's only going to be about six to nine months until some supply ends up coming back online. But who knows what happens between now and nine months from now? Like there could be an additional explosion of compute needs if you know, there's new agentic models that end up coming out another open claw like breakthrough that ends up driving more demand.
No one could have even have seen open claw coming. So there's always ways for me to turn this trade into a little bit longer of a trade. But not if it goes down. If there's a break in the thesis, I will sell and I'll even sell at a loss if it if that's what it comes to, but I don't think that's what's going to happen here, obviously.
Like everyone else, I mean, you think your trades are going to work especially on day one. So we'll see. Um Ruben uses Micron heavily? Yes, no, 100%. 100%. So I'm really excited to see what is shown off today. Are you invested in Coreweave? Not invested in Coreweave personally. Every single time that I go to buy Coreweave, I end up just buying more Nebius.
I it's a problem. I have a I have a hard time buying Coreweave over Nebius. Um let's see. Ether yet still above 2300. Congratulations to those that believed in uh crypto through that time. I mean, it it could sink way back down to 1,600 or something like this. I mean, it might not be the bottom, but we've seen some definite uh big whales start to recently come in.
Uh Eric Voorhees just recently put in $50 of his own money into Ethereum recently, and now all of a sudden it's starting to pump. So, maybe there's some additional information that the big traders are starting to see where uh the the the short are calling for a bottom here. Okay. Um Welcome aboard, Tanner.
Let's go. All the memory lads in chat giving me the welcome. I'm I appreciate you guys holding the door open for me. That's awesome. You didn't need to roll out the green carpet for me. I say green because that's just a better sign in the markets, and I'm a big Nvidia bull, so um no red carpets here.
Only green. Speaking of red and green, oil. The price is not uh low, okay? 94.55 a barrel, but uh you know, early this morning we were above $100 a barrel. So, to end up seeing that drop over this morning is very exciting to see. However, we're still at that 94.60 range, above 90s. You know. Also, we didn't even take a look at just the broader markets.
Let's go take a look at the the QQQ's. The Q's are up 1.3%. It also helps whenever the biggest company in the Nasdaq 100 is up 2.5%. That that obviously uh is nice assistance. And then, the S&P 500 is up 1.1%. Wow. Okay. Um Let's go ahead. There was also a video here that I wanted to check on CNBC regarding uh Nvidia, which I think I haven't seen this video, but I think that this is in preparation for GTC.
So, let's have a listen to this really quick. Only 4 minutes long. Welcome back. Nvidia hosting its GTC conference in San Jose this week. Kristina Partsinevelos on the ground with some of the most important announcements we should be watching for. Kristina. Well, sorry. For years, Nvidia's pitch was simple.
One chip for everything. Training AI models, running them, answering your queries. The GPU does it all. But, that pitch is now under pressure. The AI industry has shifted even just within the last 8 months. Companies aren't just building models anymore. They're running them, generating responses, powering agents that write code and answer questions.
That process is called inference, and by 2030, Bank of America estimates it could account for roughly 75% of an over $1 trillion AI market. The problem for Nvidia, though, its flagship chips aren't built for it. They consume too much energy and don't carry enough memory to efficiently spit out answers at scale.
Google, Amazon, Meta, startups like Cerebras are already all working on alternatives within the inference space. So, today in the arena right behind me that holds over 15,000 people, many of them waiting in line over 2 hours ago, look for CEO Jensen Huang to come out swinging. He's expected to unveil a new inference chip built on technology from Groq, a startup Nvidia uh paid $20 billion for Groq's design uh just within the last few months.
That design puts memory directly on the chip, making it dramatically faster and more power efficient for real-time AI tasks in inference. It's Nvidia's answer to every customer who said its GPUs were overkill. The options market is pricing in roughly a 4% move, which is modest by Nvidia standards. The real catalyst though may not necessarily be the chip announcements.
It's about demand visibility because so many on Wall Street already know about these products or guessing these products are going to come out. It's been 5 months since Nvidia told us they had a $500 billion back order at GTC DC. And Jensen Huang updates that number and gives visibility into 2027. They did already give an update on that in their last earnings.
They said they're way above the 500 billion. So maybe they can give even more clarity, but that was already cleared up in their latest earnings. That's what moves the stock. The products are definitely the show, but the numbers are the trade, guys. Christina, I'm just thinking about other things he might potentially, I guess, talk about.
One would be the impact of the war on energy exports and hence semi-fab production. Maybe maybe we try to get something on China and whether or not talks are allowing a path to make that revenue a little more visible. It seems though most sell-side analysts overall have just written off China, so it's not included in the estimates.
So if he does make a comment about China given the the I guess ease now for exporting certain AI chips to China has been lifted. The United United States government is more on board with that. That could definitely be a positive force the stock, but I think Jensen Huang's narrative is shifted towards sovereign AI all around the rest of the globe.
Uh last earnings he said it contributed roughly $30 billion. So I expect him to really focus on other countries as opposed to just China. And in regards to the war, I don't know if he's really going to chime in on, you know, whether helium is going to affect the production of their chips, but you never know.
Yeah, I I also, you know, partnerships, right? Nvidia has made it such a central part of its strategy here to both invest and partner with so many different parts of the AI ecosystem. And I wonder if there's speculation that more is on the way. Oh, 100. They have a lot of money in their pockets right now to be spending, so I guarantee that there's going to be more partnerships, uh more supply deals, working with not only, you know, competitors, working with uh every even like with recently the $2 billion uh deal with Lights and uh or Lumentum and Coherent.
That's within the optical market. So, we're expecting to hear news on how they're going to change their networking uh call for towards optical, so using light. That is a big partnership and really just shows how Nvidia is trying to move through the entire stack and use that money. And the other one quick thing, I wonder if they'll mention anything tomorrow about buybacks, because there hasn't been any announcements in the near term.
They have a lot of cash, and tomorrow the Q&A with analysts might reveal something. Yeah, Roth Research says potential collaborations including Intel. They They raised that as a possibility in the morning note. Kristina, we'll be watching. Yeah, I mean, if we end up getting a uh price update, you guys will see me uh you know, commenting in chat while we end up watching the show together.
But, if they end up showing off some new number, some updated forecast for the next uh six quarters or something like this that that highlights uh 700 billion, 800 billion dollar, who knows what the number that they're going to throw out. Um uh that is what is going to push the market. She's absolutely correct.
Uh but yeah. Nemo clause already priced in. All this stuff about priced in, priced in this, priced in that. The The stock end up growing way higher than expectations, and then the market didn't move even in the slightest. The amount that's being priced in is so not true. There's so much is being discounted, not being priced in.
Um they don't believe what's actually happening until they actually deliver on the numbers. So, while I do think that GTC could absolutely move the market and it did last uh November, overall, um Oh, wow. You already see Look at this photo. Oh my gosh. Dylan Patel of all people is seated this far away from the actual event.
Geez. Oh, speaker. He's a speaker. I guess that makes sense. That is just crazy. He's not going to be speaking at the keynote, but I'm sure he has uh an event that he will attend. Um Okay. Rip my boy Dylan. I know. He's He's He's not getting a good view. I'm curious to see where uh where you know everyone else is going to be.
Oh my gosh, Jim Cramer's posting about uh Nebius being there. Here, wait, let's see this. Nebius and Invertive Oh gosh, I think this alone is uh is bringing down some of the the AI stocks right now. Let's go see where Nebius is. I mean, that that con- continued to climb up. Still up 15% above $130 a share.
I tried to defend Nebius um yesterday on a show with The Mitten Steve and everything like this, and I don't think I got a lot of my points across the way that I wanted to, and and sometimes um you know, I I can be misinterpreted from how I'm actually feeling about these companies. But, overall for like Nebius, the way that I think about it is very simple.
I'm a total addressable market investor, okay? You want to pick companies that are early in their total addressable markets, but those markets are expanding very quickly. So, if they can get a foothold in those markets while they continue to expand, then customers are going to look to them to fulfill their needs.
And I think that Nebius has placed themselves very well there. Uh you you even look at the the Metas and the Microsofts, and that's fine. Like like the hyperscaler names are very exciting to to end up getting the stock to pop on that type of news. That's not what's getting me excited about Nebius. You know, my my excitement is I think that a lot of uh enterprises and and like all organizations are going to be using artificial intelligence.
And they're going to want like easy solutions out of the box to end up working with their own data uh in the cloud. And and they'll pay for that premium cuz they're not big enough to set up their own NVL72s, but they are big enough to end up signing multi-year contracts. Um and so they'll work with with companies like Nebius.
And I think Coreweave's going to be a winner. I think Iren's going to do well. I think, you know, all the hyperscalers, Oracle, they're all going to do well. Except there's there's small tweaks in each of these companies that I think are um that make me more comfort like feel much better about the space.
So, one, Nebius and their entire head count have been doing this for a long time through Yandex, a previous company with 20-plus years of experience. And then on top of that, you have um you know, one, they're they're completely AI cloud-native, so you don't have the uh Microsofts and the Amazons and all of these things that have traditional businesses that are also doing very well.
I'm not shorting, you know, um Whole Foods and amazon.com and all of these things but they're diluted versus how fast the cloud business is growing. Whenever you look to Google Cloud, although search is doing very well, if the high high majority of what you're buying is something that's growing at the teens percentages of growth or or Google network might be shrinking a little bit, then what you actually want is the cloud thing and that's what you're betting on.
It's just not nearly growing as fast as you know, purely Google Cloud or purely AWS or even as you go down into these smaller names like Nebius, they're growing at 500 plus percent. They just signed a 5-year deal that's essentially the size of their company in revenue. They're not going to make that all in margin for for pure pure profit, but it is profitable.
They're not signing 5-year deals to lose money over those 5 years. Okay. Um And so it was hard for me to get my my point across to um to a mitten in in Steve and these sort of people, but I just think the total addressable market is expanding so quickly and you want to get in early before you know, they end up signing those contracts.
So, if you were able to buy at the $70 prices, $80 prices whenever we were really banging the table, obviously now is a nice um you know, comfort cushion. If if you even purchased before then in the 20s, 40s, you're not even worried at all, but I I I think if you believe where this is going, where the the meta deal is not going to be the last one.
There's going to be a whole lot of customers that end up coming in to buy smaller, higher margin allocations of compute from them as well. That it'll be margin accretive, they'll continue to grow. You know, we're we're very very early on in that space. Um do you think that this runs aggressively now?
I dude, I I think this is what I told um this is what I told Emitt and Steve. Everyone's giving me credit because they signed a deal the next day. I had no clue about this. But, it's where the space is going, okay? So, I didn't know when it was going to happen, but I know that there's going to be deal signed with them and this will continue to be a creative to the stock.
Do I know like how does the market think about Nebius after the meta deal? I don't know. I know how I feel about it, okay? It's additional compute. They had a customer that came in for $3 billion and then they said, "Oh, wow. They're delivering on time. Let's 10x that deal." That is in insanely exciting, okay?
They didn't just go and sign with them all up front and let's see if they performed. They did $3 billion. They loved what they saw. They did 30 sorry, 27 billion more. And they they can dilute, but that's in order to raise for a profitable business. On top of that, they're trying to get a lot of this a lot of their deals with upfront 100% upfront payments.
So, I think that it's going to be a lot less dilution than people give it credit for, okay? Meta could pay them a hefty hefty chunk of cash in order to set up those data centers, more than people would give it credit for. Um what do you think of the Iran and Nebius beef on X? I think both I think both will thrive.
I think both will thrive, okay? However, just like whenever I'm picking Nebius over Iran or Nebius over Coreweave, I do like Nebius more. There are things that I see in Iran that just worry me. And obviously, I see what people see as well. They have a massive amount of contracted power that could be huge revenue in the future.
But, the fact that they're not actually signing it and and they have um you know, chips to spare scares the crap out of me. Why is that? Nebius doesn't have a single chip to spare. They're they're they're turning away customers because they realize like every single time that they set something up, they immediately sell it.
AWS said the same thing, Google said the same thing, uh that they're monetizing compute the second they get it online. The companies that don't share that same sentiment scares the crap out of me. But but I still think that they'll do fine because the market is growing just that fast. Okay. Um seriously, what's wrong with SoFi?
Where is SoFi right now? SoFi down 1%. I mean, honestly, they're getting hit with uh private credit, they're getting hit with the chance of lowering um the the the lowering chance of rate cuts. They're they're they're getting hit left, right, and center. But, honestly, that's why I'm mainly focused on the AI trade right now because the macro is good for the AI trade.
You know, I'm I'm I just bought Micron. I'm I'm doubling down on uh the hyperscalers. I'm doubling down on the neo clouds. I'm holding 50% of my portfolio, 50% of my portfolio is in uh Nvidia right now. Rate cuts are good for SoFi according to Gemini. A Gemini's right, but the chance of a rate cut is going down.
So, that's that's you're Gemini's 100% right. But, if we don't get that, then that then that's not good. Um Good relations with the Russian government is a huge plus. You uh you're not reading into uh Nebius these days. That's an old uh overused comment. They are European. I don't remember which state.
Do you guys remember? Are they Dutch citizens now? I I forgot the actual state that they are in, but um yeah. They're not Russians. Um when do you think the strait will be reopened? I'll stick to the things that I, you know, feel like I have any bit of strength in. Um knowing when the Strait of Hormuz reopens is not one of them.
So, while I will follow that news and and I I I'll get excited whenever it opens up because it's good for for markets in general. I I have no business trying to guess macro at all. I let the macro do its thing and um I try to figure out which companies are going to be the best for whenever they reopen.
Okay? Um However, let's listen to Scott Bessent because he's one of the guys that uh definitely does follow macro events and uh the people that I learn from. So, let's just go straight to the source. Uh here it is. I was going to ask you cuz you this is the old hedge fund manager Treasury Secretary coming out counting those barrels in your math cuz everybody has different numbers.
How many barrels of oil do you really believe your office, you personally, the president believes that we are short in the world right now? Well, it looks like the deficit is about 10 or 14. And that's before any of the the ships are coming out of the straits. So, if you think about the Russian oil, that is somewhere between 9 to 11 days uh 12 days of supply with without the market moving.
Uh there's Iranian oil that's out in the water right about 140 million barrels. The Saudis, they have supplies stored around the world that they can release. And then the global SPR release, largest ever, it was 400 million barrels. Is there any scenario which the White House or Treasury would extend that waiver on Russian oil from 30 days to something longer?
And you understand also the optics of that. We hear we're moving basically sanctions on Putin, people react to that. Well, Brent the which is which is more? The if oil spikes to $150, but Putin was getting 70% of that, or oil stays at 95 to 100, where is he getting more money? If it spikes to 150. So, roughly his coffers are unchanged.
And that it whether a couple billion more, a lot of money in the real world, but it in terms of the Russian Federation, costs him about two mi- $2 billion a day to run the government. So, uh it the idea is as soon as the conflict is over to and the markets are well supplied to go back and again, it is a short-term window for oil on the water.
We we believe that that will be that consumed very quickly. And Brent, the other important thing to know is that oil was going to China. They were going to consume it anyway. It was just backed up. Yeah. And so, by unsanctioning it, it can go to Malaysia, Singapore, India. Uh the Indians have have bought quite a bit.
Uh so, it can go to alleviate the shortage because most of the oil from the Gulf does in fact go to Asia. It does. Uh want to wrap it up. Just I have to ask this question. So, just kind of going back to my earlier question, is there something besides a global oil release that you or your office, the White House, could do.
Could you intervene in the financial side of the commodities or oil markets to bring down the price of oil? Yeah, that that rumor's in the market. When there's When you know when there's big dynamic price action, that always happens. We we haven't done that. You have Would you do it? So, the I'm not sure under what authority or what auspices we But you have now I want to be clear cuz you know that that's that that that's out there.
Don't worry, the Treasury's in the in the Plunge Protection Team is in focus. Yeah, you know, people all people used to say like the Treasury was buying S&P's. I think that's usually excuse for people who got too short and got squeezed. And then finally wrap it up I got got to ask this to wrap it up.
Um expectation of timeline on this because with every passing day, with every passing week, now we're in the third week, market's going up, fertilizers, gas, everything. And And the world is really worried about shortages and also the inflationary effect here and abroad. Yeah, Brent, can we just be clear here?
It's 2 weeks. It It's 2 weeks and you know I know the media likes to move things along. We see these headlines and that's part of the Iranian strategy and it's very unfortunate that because of a dislike for President Trump, not you, but a lot of mainstream media are trying to speed this up, trying to make it into some crisis that it's not.
And on the other side of this, if you look at longer-dated crude, you look at future inflation expectations, they're both very well anchored and you know this this will end and I don't know how how many weeks it will be, but on the other side of this, the world will be safer and we will be better supplied.
So, the backwardation of the market, you think is correct? $80 oil in a couple months? I think probably much lower. Much lower than $80 a barrel a few months from now. Okay. Um GTC is in 25 minutes. Nvidia is still looking very hot right now and we're starting to get some news reports coming out already of some commentary from Jensen.
Along with uh Jose Najarro on X ended up just tweeting this. Currently at Nvidia GTC 2026, one thing I can say for certain is the market is 100% under appreciating the amount of AI compute is needed due to this inflection point driven by AI agents. I've talked to a few software people and other tech heavy creators and they are so bullish.
It makes me want to tinker with open claw today. It's actually crazy that names like open claw and all of this stuff that is very exciting. But how is that going to be put at the enterprise level? We're not going to be buying, you know, Mac Studios on the enterprise level, so you need things that will scale into the enterprise, right?
And that's where I think things like Perplexity computer, base 44, uh Manas uh com- computer, which I think was announced today. Where is that? Um I did just have a thing about it. Did I not? Anyway, Manas announced something else which is also part of Meta. Um and also especially Nemo claw, which is Nvidia's own version of this.
Now, supposedly Neo claw Nemo claw does not need to work on um Nvidia hardware and it would work with anyone's hardware, essentially across Nvidia products, which is quite interesting. It's definitely the space that that Jensen's been really open to do. Like you don't need to buy everything from Nvidia, but you need to buy something.
Okay? So, this could be networking, this could be some of our chips, some of our switches. Like, who knows? But, you got to be a you got to be a customer of Nvidia. Which I think is great, but it'll be interesting to see what actually blows up and and becomes extremely useful. Maybe it's Neo Nemo claw.
Maybe it's their own frontier models, which are on perplexity computer and these sort of things. I'm pretty sure I actually have access to those models. Do I not? Um Where is it? Yeah, Nemo Tron 3 super is available on perplexity computer right now. Okay. So, who's getting ready for GTC? What do you guys want to uh What do you guys want to see from them?
Tanner, the demand is going to be from the new AI TAM created by those claw bots. Mac mini prices are already going up in the past week for people like me buying Mac minis. Soon average Joes will need 32 GB. I I I I'm not saying that that's not going to continue. I just think that there's also a space for enterprise level NVL 72s to contin- continue to be sold, right?
And that the future of AI agents running on the cloud is not like I'm because I said that I think that Nvidia is still going to sell products. It's not that, you know, smaller systems are not going to be sold. Of course, they are. But, so are the big chips. The individual user might not need a Vera Rubin system, but um Meta will.
Amazon will. Salesforce will. All of these names will. I want Jensen to make normies care. Carl, it's I talk to people all the time. It's mind-blow- I'm like I'm like, "Oh, have you been keeping up with these models?" And they're like, "Ah, I don't use AI. It's it hallucinates and all this stuff." I was like, honestly, for what you use it for, nobody even it does not.
Like, if you're using it for basic questions and answers basic People think it's still like bugs out with math. I'm like, if you put on a reasoning model, everyone thinks that it just predicts the next character. But with reasoning models, it finishes the prompt and then looks it over 50 separate times and then realizes, "Oh, it got the answer wrong.
Okay, let's rewrite the real mathematical question." Or the real math- mathematical answer. Like, the the hallucinations are getting lower and lower until you start dealing with uh you know, prompts that take hours. Prompts that take extreme amounts of token usage. Then you might end up seeing some slippage, some things that you didn't really uh you you didn't want didn't connect to tools perfectly, but that's the new area where we need to solve those problems because the usage continues to expand.
And then so I I I I talk with these people online. I'm like, or not online, like in in my you know, real life. And I'm like, "Oh, how many Like like have you seen any of these agent stuff? Have you seen Perplexity computer base 44? Have you seen what Cursor does? What what Claude co-work does?" No's across the board.
People haven't even seen it. They haven't seen an update of it. It's crazy how nascent we are right now. Like, yeah. I essentially am and and this is a comment from Jensen that just appeared, by the way. Jensen said that um AI agents are going to make humans more useful, not less useful. His point was that agents will handle the boring repetitive work, while people spend more time on higher level of thinking as projects and products get delivered faster.
And so, uh for for people that are new to the channel or whatever, I'm hiring on uh my first employee for future investing at the end of this month. And I was talking to them about a project that I wanted them to work on pretty soon. I don't want to give that away, but essentially, I let them know this.
And this is a uh more on the tech savvy side of people. And he was like, "Oh, it's going to take me a bit to uh be able to create that, but we can definitely, you know, work on this." And I was like, "Honestly, I think that you can be able to do this in a few prompts on base 44. Like, it's not that complicated.
Like, don't overthink it. I know it's a big project, but honestly, where AI is right now, this is super doable." And just to show, like, just to show uh the basics of what I wanted, I did one prompt, put in what I wanted for this tool to do, in base 44, it spit out the exact thing that I wanted. Now, it was super, super cut and dry in terms of like, um the amount of material that it actually had was way, way less than what I actually wanted it to have, but it was one prompt.
And I told him, I said I said, "I made this in 3 minutes. Don't freak out. Like, you'll be able to make this with a few dozen prompts, exactly what I want you to make. And the reason why I'm hiring this person is because I can't I can't focus on all the major step-ups in artificial intelligence, but the brand needs me to, you know?
I'll I'll fall behind if I don't, but also, I want to make sure that I'm keeping up with stock content and making sure that my trades are good and and all of this stuff. So, I'm not focusing wholeheartedly on every new AI agent and and solution that ends up coming out. So, I'm just going to pay a person full-time to do that.
Um And And the reason is is because, well, I think that they're going to deliver a massive return, and I think anyone can be able to do that. I don't think you need to be a prompt engineer in order to realize Like the way that these things work is it works off natural language processing, right? Like it's really going to to remove the barrier between coding and and what I do, essentially just talking.
So, if you're the create like that I think what made people coders and creators to begin or a coders and developers to begin with is they love to build things. But, one of the one of the processes of building things, especially in software, is that translation between English or whatever the language is into, you know, Python or what whatever the language that you choose, but but essentially into code, okay?
And And that language barrier is what stops people from being software developers. These new AI tools are the translators. So, you just need to be the building type, the creative type, the the people with ideas, and those are the people who are going to excel into the future. And And that's going to be the big difference.
The ideas are going to separate people, not the ability to translate the language. That's super interesting. Now, to be fair, a lot of the um you know, big advancements in how Spotify ended up making their solution, how Facebook ended up making their solutions, uh they essentially had to reinvent the way that code ended up working in order to make their solutions work.
Like there are nothing There There had been nothing done the way that uh a Spotify or or even a Shopify or or a um Facebook, they had to sort of be extremely creative. I don't know if bots will be able to do that because essentially what they're grabbing is all the historical nature of of code and then end up creating something.
At least for now, that might be a barrier. So, what I noticed from a lot of people's codes that they are doing right now, and this will be solved eventually, but just right now, um you see everyone coming out with all these passion projects and like like their their vibe coding uh projects into the world, but they're all extremely clunky and slow because the the the AI agents are not optimizing for speed and latency and less code.
They're not trying to tighten it. They're just writing a ton of lines of code. It actually makes the a lot of these websites that I'm checking out really slow and clunky and a little bit ugly, in my opinion. The yappers will be the better coders. I promise you guys. Okay. That being said, uh 15 minutes until GTC.
Let's go have a look at the markets and get ready for uh GTC. So, first up, Nvidia up 2.3% 184. We're going to keep an eye on this. I'll be typing into chat uh and throwing it up on screen what the price is as we go along. Okay? So, we'll continue to update that. On top of that, we're also going to keep up with how other names are moving like Nebius, like Meta, like Amazon.
Some of these names um that could end up showing massive partnerships. Microsoft, I think is uh almost certainly going to get uh a shoutout here for their Vera Rubin platform. Um they got the first Vera Rubin NVL 72 for validation up. We can uh Satya Nadella. Where did Satya Nadella post this here? So, I'm sure this will get highlighted.
Um We're the first cloud to bring up an NVL 72 Vera Rubin and NVL 72 system for validation. Another big step in building the next generation of AI infrastructure with Nvidia. You got to love it. Absolutely beautiful. Absolutely stunning. Faster to set up. Faster to produce. Uh it's you know, it's tubeless.
It's it's essentially cordless. There's not miles and miles of of copper cabling and everything. So, uh much faster to produce, much faster to deploy. Meaning more customers can come back after they see an actual a real revenue from these, right? Whether that's token generation, whether that's training, what have you.
But, um there there's going to be real return on Vera Rubin much faster than we ended up seeing uh Hopper, Ampere, Blackwell, or any of the others. Okay, so this is very very exciting for um Nvidia. And that's not even including what we're about to see him talk about in this update, which supposedly has to lot has a lot to do with uh Grok, their recent uh licensing agreement.
I think it was like a $20 billion licensing agreement essentially. I'm an AI coder for 10 plus years. No chance AI will replace coders. Okay. Um I want to short AMD. I would not do that either. There there's going to there's going to be a lot to go around. Um obviously, I'm a big believer in Nvidia doing quite well, but Um Office 365 agentic is in pipeline and I already have it working for me.
I love seeing whenever you guys um are using the AI agents. Maybe there's more of you in this chat cuz we're literally all about to watch Nvidia GTC together. But, boy oh boy. I'm pretty sure the event, by the way, is about 2 hours long. If it is exactly 2 hours, and that leads us from 2:00 p.m. to 4:00 p.m.
where the market ends up closing, I'll just go directly into a market close to watch what Nvidia does after hours. Okay? In terms of companies that are showing off earnings today, uh Semtech, I don't really care. I don't really care about Semtech. Tomorrow we have Oaklo, DocuSign, Lululemon uh after hours.
That might be interesting. Wednesday we have Micron, DLocal, Redcat. I'll be listening to Micron's earnings live. Even though I have a much larger share in DLocal, I think Micron's going to be uh much more important for this overall AI trade. Okay? I ended up just purchasing some uh Micron this morning as well, so I have a trade for earnings.
I'm doing an earnings trade on that name specifically for a massive guidance increase. If it goes wrong, hey, I mean, it's another indicator that I shouldn't trade on earnings, but whatever. We will find out. Okay? 10 minutes to go. 10 minutes to go. Let's check on gold and silver really quick. I haven't checked the the commodities today.
Uh gold is down 0.6% and silver is flat on the day, 72.68 an ounce according to uh the silver trust ETF. Uh Ethereum up to $2.3,000 uh per coin and Bitcoin up to nearly 74,000, 70 uh 3 800. And we actually just saw so uh ETHA, which is iShares Ethereum Trust, which ended up ended up ballooning up to uh uh 10 billion dollars of AUM, since shrunk down since Ethereum has sold off so much.
But, uh this fund only was uh essentially adjusting for the price of Ethereum, while also having uh wow, what a candle. While also having no ability to stake those rewards. So, then now we ended up just seeing the publishing of ETHB, which has no AUM right now, 150 million. But, I think that this one's going to be much much bigger.
Uh ETHB is their staking Ethereum ETF. Meaning that while you do have to pay um you know, a management expense ratio or MER for this fund, they also have staking, so it essentially pays for itself. The other one, you still have to pay for a sponsorship fee or an MER, but there's no way to make passive income over time.
The only reason why you would end up using these rather than just buying direct Ethereum, at least for me, would be retirement accounts, stuff like this. Um to potentially give you allocation into an account that might have uh sort of stopping you from from buying anything other than equities. It essentially is an equity, but it's buying a crypto.
All right. Uh Bitmain, by the way, also another name here buying uh Ethereum is benefiting from the recent price increase. TeraWulf up 12.4% SanDisk up 7 uh 7.7% NAND is obviously going to be uh really big here. And then we also have uh Micron up four, nearly 5% right now. Awesome. Okay. Let's see what's not moving today.
Well, supposedly SoFi is not moving, but let's see what else is not moving. Better homes, the VIX is down a ton, recent uh recently public Pay Pay down 7%, VG is down 6.4%, Klarna down 5%, Fastly, Gemini, Rubric. Surprised Gemini's down 5% on uh crypto being up here today. Cloudflare down 10%, CrowdStrike down 3.2%.
Very interesting. Okay. The markets are green. Let's not focus on the on the red here. Circle up eight, SanDisk up 7.7, Air Test Systems up 7%. It's been like one of the most volatile names that I've I've followed is Air Test Systems. Um If you only had $100 to invest, what would you do? I would uh buy some nice clothes to buy like a suit or a razor or something, clean myself up and get a job.
Don't worry about what you're investing with $100. It's not going to make a damn difference. Go buy the QQQ's and then get back to work. It You you need to focus on whatever you can do to make the most amount of money, okay? So, if that's your income right now, focus on that uh your ability to produce more income, and then once you have more money uh to invest, then worry about uh the stock markets.
If you only have $100, man, don't invest that. Uh yeah. Do you work a job, Tanner? No. You're watching me do it. This is uh this is the job now. It was just a thought experiment. Uh I mean, pick a thousand dollars and pick If you could only buy one stock, what would you buy? I don't think that I I could I don't know if I could pick a name.
Nvidia? I don't know. Maybe I'd want some more diversification, maybe Amazon. That might be If I could pick only one name, Amazon seems like a a great pick. Seems like any amount of innovation uh Amazon would be the name to benefit from. You know, whether that's uh stablecoin innovation, self-driving, robotics, um you know, more cloud compute and and better AI models and uh any amount of of innovation I feel like Amazon is well placed to benefit from.
And because of the amount of warehouses and overall just space that they have around the world, that is so hard to replicate in anything less than a decade. So, if they had any strong competition coming for any of their businesses, you would see it coming from a mile a mile away. Are we going to watch the keynote for uh here?
1,000%. I've got the the the the keynote ready. We will be enjoying the the the the sweet words of the great and powerful Jensen Huang. All right. But, that's in 5 minutes. But, we'll absolutely be live. Um and as But Saul here says, the great thing about investing is you don't have to pick one. A little bit of diversification goes a long way.
Let's uh check out Nvidia here. However, if you are only investing with a hundred dollars, you might not have uh you might not like Honestly, as you're getting started and you're just looking for a savings account almost, I would I would worry about investing at all, but if you do want access to American markets, I would probably just buy the QQQ's and and study up during that time.
Um Are we taking a shot every time Jensen makes a joke? Uh I'll be blacked out by the end of that. I don't want to take on that type of risk. Uh will Jensen pump SoFi? I I can't imagine so. Will he pump the memory names? Will he pump uh the Neo cloud names? Will he pump the hyperscalers? Will he pump Meta?
Almost certainly. Nate in chat, the great and powerful oracle of Nebius, says DA Davidson just put a $200 price target from 150. He doesn't even have to say Nebius. I know what he's talking about. He doesn't even have to. Okay, guys. Let's get ready for the uh the event. Nvidia GTC will be live. And uh it'll jump over here in a minute.
There is millions of people watching around the world, tens of thousands of people there in person. This is going to be super cool. I can't wait. And this is the this is the big one, right? So, uh they obviously do a lot of events, GTC sort of at the end of the year, uh this big one here at the you know, headquarters.
Um They have uh CES. They do a lot of things, so be very exciting uh to see what they show off. Obviously, their main data center products is what analysts are going to be looking forward to, as well as any uh price change adjustments, them talking about, you know, the how they're handling supply constraints, whether that's the chips are more energy efficient, less memory demand, even though we all know that's not going to be the case.
Um all of these things are going to be you know, very important to pay attention to. How many customers are lining up to buy? That'll be ex Okay. Your friend Emmet is there. Yeah, no, I I I think every Nvidia creator so got invited. Um I just didn't want to go. I don't like I honestly don't like traveling.
So, we're staying home. Um Leather jacket, anyone? Oh, I forgot my leather jacket. Uh I could probably go run and get it for the for the good luck charm. Yeah, I probably should have brought a leather jacket. Okay. We also got a pretty big update from Tesla's Terafab, which is expected to start here in the next few days.
Looking to design their own chips. It'll be interesting seeing Elon put his spin on things. Right now, they're essentially aiming, this is the this is the target aspirations for Tesla's Terafab, is to target 70% of TSMC's global output. 2-nanometer chips built in-house trying to target 1/10 the cost of an Nvidia Blackwell chip at 1/3 the power consumption, and they want to do this with a $20 capex budget, and it's launching next week.
I think that's going to be hard to do. But, I'm not betting against Elon, that's for sure. Okay. Guys, I'm on the event. Okay, so so just so we're aware, this is what I'm seeing right now. We're just sort of uh waiting. So, be mindful. I'm I I'm on the right thing. Oh. Is this a render or is this a camera?
It's probably a rendering. It's Nvidia, for God's sake. But gosh, it looks nice. Jose is there, Amidst there. Yeah, they're all there. Um but I'm pretty sure they're just playing some music right now. I I I can't play the music for very long, so I'll just flash it really quick. Yeah, okay. So, they're just playing music.
But once we end up switching screen, I will uh switch. And then if there is an actual video or something like that, I'll I'll play the audio. I'm not worried about watching the thing. Okay, I will keep you guys updated on the actual um on the actual price as time goes on. I'll sort of just put something in chat like uh Nvidia stock up you know, 2.5% to uh 184.
Something like this, I don't know. Uh and then flash it on screen. So, as we're watching, I'll keep you guys updated with price chart. And I'll I'll use the periods rather than the commas next time. Okay. Still playing music. No jacket, no views. I'm dude, I've I will get the jacket on. You got to give me a second.
I'm just waiting for the the video to start playing. I don't want to leave you guys in the quiet uh quiet thing here. Oh man, everyone's at GTC right now. I saw ticker symbol U there, Mit, Jose. I don't think Neil's there. Um Yeah, there's a lot of very exciting uh creators that are very bullish on the AI space.
Now, I'm pretty bullish. I think uh Jose might be up there on on who's more bullish than me, but uh I don't know. I honestly can't foresee myself buying anything other than artificial intelligence companies right now. Like I said, I bought in uh Micron right before this. I think that we're going to see some uh talk about it here for Vera Rubin Ultra, and then we're also going to see their earnings on Wednesday, and I want to play those earnings badly.
Um Yeah. Okay. Still playing music. Uh Has anyone ever told you that you remind them of a white Urkel? Do you know who Urkel is? I think Jose is there. Well, I appreciate that. I have heard that before, but let's just I just wear glasses, guys. I got the Nvidia cap on. Calling me Urkel out here. Like let's I just have glasses on.
That's it. Okay. Uh buying Micron at the top. You are brave, sir. Thank you. Thank you. I think whenever a company is expanding earnings as fast as they are, I think like uh you know, buying the quote-unquote dip is not the strategy. Like the the the the top of the can or or the top of the chart is not where I believe that their height can be.
Right? I'm looking at the valuation. I'm looking at how fast they're expanding earnings. I'm looking at the supply constraint. I'm looking at the potential for for costs on chips to go up two three times X from here. I'm looking at um you know, they're they're acquiring multiple companies to get clean rooms online.
Like they they are essentially they just bought uh 570,000 square feet of of fab space. But it's not going to come into production till fiscal 2028. What do we do until then? You know? Nothing's going to come online for another 6 to 9 more months across SK Hynix, Samsung, Micron, all of these names.
And yet the demand for chips is only going to decline. So so price is expected to double or triple. So yes, it is one of those strategies for me that I'm going to buy high and sell higher. That's the that's the hope. Still hearing music. Um Now why are they late? Let me double check their Okay, I just checked their main website.
It's still It's still playing the same thing that we're seeing here. All right. They are slightly late, ladies and gentlemen. Um Did you sell SoFi to buy Micron? No. No. Just earning a little extra money doing some covered calls here and there. Allow me to make uh 20 minutes late? No, they're 5 minutes late right now.
Okay. Only 5 minutes late. Um just like their earnings, I'm seeing a pattern. They were late on her They were late on earnings about 15 minutes. So, let's hope that they do better than that. But, this seems like a much bigger event to make sure that everything's working fine. I remember CES went wrong, right?
They tried to put up or switch a bunch of slides, and then they didn't end up uh the slides didn't end up coming up. And essentially, Jensen had to improv everything. But, that was done at uh the Fontainebleau Hotel in Las Vegas, not at their headquarters. So, I imagine things are going to go a little bit smoother considering they've probably been prepping for I mean, freaking months for this.
This is the This is the epicenter of um of AI right now. Um Jensen was low-key pissed. Oh, dude, he was. He He was putting on a nice smile, but um he was a little little pissed. Okay. The And Nvidia is seemingly seeing some green candles right now. Um even though it just says their event will begin shortly.
Yeah, still still listening to music. But, we're up 2.6% not not not as of 2:00, but uh as of the price of the full day up 2.6% to 184 99 on the stock right now. Hope for a a close above 186. That's seeming seemingly the target right now. I'll keep you guys updated as we watch. Um Can you review some of your AI stocks?
I I mean, right now, I'm honestly like 50% in Nvidia. And then like I said, I just bought a little bit of Micron, but a lot of it is in the major names, too. Google, Meta, Amazon, Microsoft, hold all of those. Um I have uh Nebius. What else do I have? Um I'd have to think about it. I don't hold too too many.
I'll go open up the portfolio. I don't hold TSM, unfortunately, just because I'm a a huge believer in Nvidia's ability to charge an appropriate margin. I said Meta, correct? I consider them a an AI name. This is the previous Okay. Why is GTC not starting? I'm not sure. Maybe a technical check. Um Yeah.
I doubt Nvidia hits 200 off of this. If they do, I'll be shocked. I mean, it would be a long way to go in a single day. However, GTC last time was where we saw the pump to 212. So, is it out of the possibility? No, but it would need to be quite a significant day. Someone said it's started. Let me go ahead and I mean I'm I'm live on their thing.
No, it's not started. Someone get that guy. Attack. Um What's Nvidia's target for this year? So, my goal is I just need to get the stock above 250 for the year. Not not that I have control. It's not that I can get it above 250, but that's the the number that I'm hoping for. Um and that's on a bad year.
Like I mean, if all went well, I honestly think this is over a $300 stock. Which I'm fully aware of what that means for the market cap. Okay. Um but let's go see what Well, this is coming up. What analysts have it priced at and then uh hopefully we end up seeing some analyst price target upgrades after GTC.
Okay, we're still waiting. They are late, ladies and gentlemen. We're late by about 10 minutes right now. Nvidia's current forecast on their stock is $274 a share, up 47.8% upside. That's Wall Street's expectation. That's crazy. 35 analysts say that this is a strong buy with one person saying that it's a hold.
And we know that that's a um J. Gold Goldberg? Yeah. Um Tanner's always late because he wants to be like Jensen. That's funny. Yeah, Barry. Yeah. Barry's not technically a uh a Wall Street analyst of this stock. However, of course, he does have his own beliefs on the name, but everyone in the industry has called him an idiot and then moved on.
Still playing music. My portfolio is 50% Nebius, 20% BM and R. Today was a wild day. It's not over, my friend. Nebius is up 15%, BM and R up 13.2%. Congratulations. And then what's the other uh 30% that you're holding? Okay. I'm 100% Nebius. Man, you have uh That's an iron stomach right there. You need some absolute cojones to do that.
But uh when will we get Robinhood gold cards in Canada? No, your your guess is as mine. Okay. Is anyone here full porting Nvidia? Obviously, we're about to go into GTC, but I'm curious. Actually, let's do a poll on that right now. There's a new poll feature on um there's a new poll feature on StreamYard.
I don't know if this is actually like I wonder how this works. Um will Nvidia pump or dump on this event? I've never used this feature. Usually, I do it in uh YouTube. So, if it doesn't do well or something like this, let me know. Type one or two in chat. That is so cool. Type one or twos in chat. One for pump, two for dump.
That's phenomenal. So, that integrates the Twitch chat and YouTube all into one. That's just great. I'm sure if you type anything other like one, two, or something like that or 11111, it's probably not going to work. Probably going to be very character specific. Still playing music. They're They are late.
We're We're 14 minutes late. And the stock is up 2.77%. I'll I'll I'll I'll make sure I keep up the event really quick, but I just want you guys to see what the stock is doing right now. Uh stock is pushing to the highest prices that we've seen on the day, up 2.76%. Okay, and a massive amount of you are saying that uh you know, we're we're we're going to see good earnings here.
A pump going into this event after already being up 2.7% does make it harder. I don't know if it's one user or one user per vote. Seems like it just because of how many people that we're getting here. Okay, cool. I can put it where I want. Very, very cool. Oh, here we go. Here we go, ladies and gentlemen.
Everyone, good luck. Enjoy the event, and I'll see you guys soon. I'll keep you guys updated on the price. Holy crap. And the the audio is on. They're not playing anything right now, but um Okay, the the wait makes it even more exciting. Do we jump into that robotics and self-driving video? Enjoy, guys.
This is how intelligence is made. A new kind of factory. Generator of tokens. The building blocks of AI. Tokens have opened a new frontier. Turning data into knowledge. And drawing on all we have learned. Tokens are harnessing a new wave of clean energy. And unlocking the secrets of the stars. In virtual worlds, they help robots learn.
And in the physical world, perfect. Forging new paths. And clearing the way for a bountiful harvest. In the moments that matter, tokens are already there. And in the miles between, they never stop. They work where human hands cannot. So, we may all breathe easier. And the smallest hearts beat stronger.
Tokens are helping us break new ground. On a scale never attempted to empower the world. So, we can reach Star Cloud 1, separation confirmed. well beyond it. Together, we take the next great leap into a bright new future. Built for all mankind. And here is where it all begins. Welcome to the stage, NVIDIA founder and CEO, Jensen Huang.
Welcome to GTC. I just want to remind you this is a tech conference. All these people lining up so early in the morning, all of you in here, it's great to see you. GTC. GTC, we're going to talk about technology. We're going to talk about platforms. NVIDIA has three platforms. You think that we mostly talk about one of them.
It's related to CUDA X. Our systems is another platform, and now we have a new platform called AI factories. We're going to talk about all of them. And most importantly, we're going to talk about ecosystems. But before I start, let me thank our pre-game show hosts. I thought they did a great job. Sarah Guo of Conviction, Alfred Lin, Sequoia Capital, NVIDIA's first venture capitalist, Gavin Baker, NVIDIA's first major institutional investor.
These three people are deep in technology, deep in what's going on, and of course, they have just a really broad reach of technology ecosystem. And then of course, all of the VIPs that I hand selected to join us today, all-star team. I want to thank all of you for that. I also want to thank all the companies that are here.
NVIDIA, as you know, is a platform company. We have technology, we have our platforms, we have rich ecosystem. And today, there are probably 100% of the 100 trillion dollars of industry here. 450 companies sponsored this event. I want to thank you. 1,000 technical sessions, 2,000 speakers. This is This conference is going to cover every single layer of the five-layer cake of artificial intelligence.
From land power and shell to infrastructure, to chips, to the platforms, the models, and of course the most important and ultimately what's going to take get this industry taken off is all of the applications. Well, it all began it all began here. This is the 20th anniversary of CUDA. We've been working on CUDA for 20 years.
For 20 years we've been dedicated to this architecture. This revolutionary invention, SIMT, single instruction, multi-threaded, writing scalar code could spawn off into multi-threaded application, much much easier to program than SIMD. We recently added tiles so that we could help people program tensor cores and the structures of mathematics that are so foundational to artificial intelligence today.
Thousands of tools and compilers and frameworks and libraries. In open source, there's a couple of hundred thousand public projects. CUDA literally is integrated into every single ecosystem. This chart basically describes 100% of Nvidia's strategies. You've been watching me talk about this slide from the very beginning.
And ultimately, the single hardest thing to achieve is the thing on the bottom, installed base. It has taken us 20 years to now have built up hundreds of millions of GPUs and computing systems around the world that run CUDA. We are in every cloud, we're in every computer company. We serve just about every single industry.
The installed base of CUDA is the reason why the flywheel is accelerating. The installed base is what attracts developers, who then creates new algorithms that achieves a breakthrough. For example, deep learning. There are so many others. Those breakthroughs leads to entirely new markets, which builds new ecosystems around them with other companies that join, which creates a larger installed base.
This flywheel, this flywheel is now accelerating. The number of downloads of Nvidia libraries is incredibly accelerating. It's at a very large scale and growing faster than ever. This flywheel is what makes this computing platform able to sustain so much applications, so many new breakthroughs, but most importantly, it also enables these infrastructures to have extraordinarily useful life.
And the reason for that is very obvious. There's so many applications that you can run on Nvidia CUDA. We support the entire every single phase of the AI life cycle. We address every single data processing platform. We accelerate scientific principle solvers of all different kinds. And so the application reach is so great that once you install Nvidia GPUs, the useful life of it is incredibly high.
It is also one of the reasons why Ampere that we shipped them some 6 years ago, the pricing of Ampere in the cloud is going up. And so all of that is made possible fundamentally because the install base is high, the flywheel is high, the developer reach is great. And when all of that happens and we continuously update our software the computing cost declines.
The combination of accelerated computing speeding up applications tremendously. Meanwhile, as we continue to nurture and continue to update software over its life not only do you get the first-time pop, you get the continuous cost reduction of accelerated computing over time. And we're willing to nurture, willing to support every single one of these GPUs in the world because they're all architecturally compatible.
We're willing to do so because the install base is so large. If we release a new optimization, it benefits millions. This applies to everybody in the world. This combination of dynamics is what makes the Nvidia architecture expand its reach, accelerating its growth, at the same time driving down computing cost, which ultimately encourages new growth.
So, CUDA is at the center of it. But our journey to CUDA actually started 25 years ago. GeForce I know how many of you grew up with GeForce. GeForce is Nvidia's greatest marketing campaign. We attract future customers starting long before you could afford to pay for it yourself. Your parents paid. Your parents paid.
Your parents Your parents paid for you to be in video customers. And every single year, they paid up year after year after year until someday you became an amazing computer scientist and became a proper customer. A proper developer. But this is This is the house that GeForce made. 25 years ago, we started our journey which led to CUDA.
25 years ago, we invented the programmable shader. A perfectly unobvious invention to make an accelerator programmable. The world's first programmable accelerator, the pixel shader. 25 years ago, that led us to explore further and further. 20 years later, 5 years later, the invention of CUDA. One of the biggest investments that we made.
And we couldn't afford it at the time. And it consumed the vast majority of our company's profits. Was to take CUDA on the backs of GeForce to every single computer. We dedicated ourselves to create this platform because we felt so much We felt so strongly about its potential. But ultimately, the company's dedication to it, despite the hardships in the beginning, believing in every single day for thir- for 13 generations of 20 years, we now have CUDA installed everywhere.
The pixel shader led to, of course, the revolution of GeForce. And then, 10 years ago, we introduced about 10 years ago, what is it? 8 years ago? We introduced RTX. A complete redesign of our architecture for the modern era of computer graphics. GeForce brought CUDA to the world. GeForce therefore enabled Alex Krizhevsky and Ilya Sutskever and Jeff Hinton, Andrew Ng, and so many others to discover that the GPU could be their friend in accelerating deep learning.
It started the big bang of AI. 10 years ago we decided that we would fuse programmable shading and introduce two new ideas. Ray tracing, hardware ray tracing, which is incredibly hard to do, and a new idea at the time, imagine about 10 years ago we thought that AI would revolutionize computer graphics.
Just as GeForce brought AI to the world, AI is now going to go back and revolutionize how computer graphics is done all together. Well, today I'm going to show you something of the future. This is our next-generation of graphics technology. We call it neural rendering, the fusion the fusion of 3D graphics and artificial intelligence.
This is DLSS 5. Take a look at it. Woo! Isn't that incredible? Computer graphics comes to life. Now, what did we do? We fused controllable 3D graphics, the ground truth of virtual worlds, the structured data. Remember this word, the structured data of virtual worlds, of generated worlds. We combine 3D graphics, structured data with generative AI, probabilistic computing.
One of them is completely predictive, the other one probabilistic, yet highly realistic. We combine these two ideas, combine these two ideas, control through structured data, controlled perfectly, and yet generating at the same time. And as a result, the content is beautiful, amazing, as well as controllable.
This concept of fusing structured information and generative AI will repeat itself in one industry after another industry after another industry. Structured data is the foundation of trustworthy AI. Well, this is going to scare you a little bit. I'm going to flip the slide, and don't gasp. So, we're going to go through the schematic for the rest of the time.
This is my best slide. Every time I I ask my I ask the team, "What's my best slide?" Repeatedly, this was it. They say, "Don't do it, Jensen. Don't do it." I said, "No. This These seats are free for some of you. So, this is your price of admission. So, this is This is structured data. You've heard of it.
SQL, Spark, Pandas, Velox, some of these really, really important, very large platforms. Snow Snow Snowflake, Databricks, EMR, Amazon EMR, um Azure Fabric, Google Cloud, BigQuery. All of these platforms are processing data frames. These data frames are giant spreadsheets, and they hold all of life's information.
This is the structured data, the ground truth of business. This is the ground truth of enterprise computing. Well, now we're going to have AI use structured data, and we better accelerate the living daylights out of it. It used to be okay, and we would, you know, of course, we would accelerate uh structured data so that we could do more, we could do it more cheaply, we could do it more frequently per day, and keep the company running at a much more synchronized way.
However, in the future, what's going to happen is these data structures are going to be used by AI, and AI is going to be much, much faster than us. Future agents are going to use structured databases as well. And then of course, the unstructured database, the generative database. This database is represents the vast majority of the world.
Vector databases, unstructured data, PDFs, videos, speeches, all of the world's information, about 90% of what's generated every single year is unstructured data. Until now, this data has been completely useless to the world. We read it, we put it into our file system, and that's it. Unfortunately, we can't query it.
We can't search for it. It's hard to do that. And the reason for that is because there's no easy indexing of unstructured data. You have to understand its meaning, its purpose. And so now we have AI do that. Just as AI was able to solve multi-modality perception you can and understanding, you can use that same technology multi-modality perception and understanding to go read a PDF, to understand its meaning.
And from that meaning embedded into a larger structure that we can search into, we can query into. Nvidia created two foundational libraries. Just like we created RTx for 3D graphics, we created cuDF for data frames, structured data. We created cuVS for vector stores, semantic data, unstructured data, AI data.
These two platforms are going to be two of the most important platforms in the future. Super excited to see its adoption throughout the network, this complicated network of the world's data processing systems. And the reason for that is because data processing has been around a long time. And therefore, so many different companies and platforms and services, it has taken us a long time to integrate deeply into this ecosystem.
I'm super proud of the work that we're doing here. And then today, we're announcing several of them. IBM, the inventor of SQL, one of the most important domain-specific languages of all kind of all time, is accelerating Watson X data with cuDF. Let's take a look at it. 60 years ago, IBM introduced the System 360, the first modern platform for general-purpose computing, launching the computing era.
Then SQL, a declarative language to query data without requiring the computer to be instructed step-by-step. And the data warehouse. Each the foundations of modern enterprise computing. Today, IBM and NVIDIA are reinventing data processing for the era of AI by accelerating IBM Watson x.data SQL engines with NVIDIA GPU computing libraries.
Data is the ground truth that gives AI context and meaning. AI needs rapid access to massive data sets. Today's CPU data processing systems can't keep up. Nestlé makes thousands of supply chain decisions every day. Their order to cash data mart aggregates every supply, order, and delivery event across global operations in 185 countries.
On CPUs, Nestlé refreshed the data mart a few times a day. With accelerated Watson x.data running on NVIDIA GPUs, Nestlé can run the same workload five times faster at 83% lower cost. The next computing platform has arrived. Accelerated computing for the era of AI. NVIDIA accelerates data processing in the cloud.
We also accelerate data processing on prem. As you know, Dell is the world leading computer systems maker, and they also are one of the world's leading storage providers. And they worked with us to create the Dell AI data platform that integrates QDF and QVS to create an accelerated data platform well, for the era of AI.
And this is an example of what they did with NTT data. Huge speed up. This is cloud Google Cloud and Google Cloud as you know, we've been working with Google Cloud for a very long time. We accelerate Google's Vertex AI. We now accelerate BigQuery. Really important framework and really important platform.
And this is an example of our work together with Snapchat where we reduced their cost of computing by nearly 80%. When you accelerate data processing, when you accelerate computing, you get the benefit of speed, you get the benefit of scale. But most importantly, you also get the benefit of cost. And so all of those come together as one.
It was originally called Moore's Law. Moore's Law was about getting performance doubling every couple of years. It's another way of saying, so long as the price remains about the same and most computers remained about the same, you're also getting twice the performance every year or you're reducing the cost of computing every single year.
Well, Moore's Law has run out of steam. We need a new approach. Accelerated computing allows us to take these giant leaps forward. And as you will see later, because we continue to optimize the algorithms and video is an algorithm company. As we continue to optimize the algorithms and because our our reach is so large and our installed base is so large, we can reduce the computing cost, increasing the scale, increasing the speed for everybody continuously.
This is Google Cloud. You could see this pattern I just mentioned. I just wanted to show you three versions of it. Nvidia built the accelerated computing platform. Has a bunch of libraries on top. I gave you three examples. RTX is one of them. cuDF is another. cuVS and we'll show you a few more. These libraries sit on top of our platform.
But ultimately, we integrate into the world's cloud services, into the world's OEMs, and together, and other platforms that I'll show you, together, we're able to reach the world. This pattern, NVIDIA, Google Cloud, Snapchat, will repeat over and over again, and it kind of looks like this. And so this is one example, NVIDIA with Google Cloud.
We accelerate Vertex AI, we accelerate BigQuery. We accelerate We're I'm super proud of the work that we've done with JAX and XLA. We are incredible on PyTorch. We're the only accelerator in the world that's incredible on PyTorch and incredible on JAX and XLA. And the customers that we support, the Base Tens, the CrowdStrikes, Puma, Salesforce, they're not our customers, but their customers, developers of ours that we've integrated the NVIDIA technologies into, that we can then land on the clouds.
Our relationship with cloud service providers are essentially us bringing customers to them. We integrate our libraries, we accelerate workloads, and we land those customers in the clouds. And so, as you could see, most of our cloud service providers love working with us, and um they're always asking us to land the next customer on their cloud, and I just want to let you know, there are a lot of customers.
We're going to accelerate everybody. And so there'll be lots and lots of customers we'll be able to land in your cloud. Just be patient with us. And so this is Google Cloud. This is AWS. We've been working with AWS a long time. And one of the areas One of the One of the things I'm super excited about this year is we're going to bring Open AI to AWS.
And so it's going to drive enormous consumption of cloud computing at AWS. It's going to expand the reach and expand the compute of open AI. And as you know, they are completely compute constrained. And so, AWS, we accelerate EMR, we accelerate SageMaker, we accelerate Bedrock. Nvidia's integrated really deeply into AWS.
They were our first cloud partner. Microsoft Azure. Nvidia's A100 supercomputer um was the the first one we built was for Nvidia. The first one we installed was at Azure. And that led to the the the big successful partnership with Open AI. But we've been working with Azure for quite a long time. We accelerate Azure cloud.
Now it's uh their AI foundry we partner deeply with. We accelerate Bing search. We work with them on Azure regions. This is one of the areas that is incredibly important as we continue to expand AI throughout the world. One of the capabilities that we offer is confidential computing. That in confidential computing, you want to make sure that even the operator cannot see your data.
Even the operator cannot touch or see your models. Confidential computing Nvidia's GPUs is the first ones in the world to do that. It's now able to support confidential computing and protected deployment of these very valuable open AI models and and Anthropic models throughout clouds and different regions and all because of our confidential computing.
Confidential computing super important. And here's an example where we have different customers that we work with. Synopsys, a great partner of ours. We're accelerating all of their EDA and CAU workflows. And then we landed at Microsoft Azure. We were Oracle's first AI customer. Most people would have thought we were their first supplier.
We were their first supplier also. But we were their first AI customer. I'm quite proud of the fact that I explained AI clouds to Oracle for the first time. And we were their first customer. Since then, they've really taken off. We've landed a whole bunch of our partners there. Court, Cohere, and Fireworks, and of course, very famously, OpenAI.
A great partnership with Court Court uh Coreweave. They're the world's first AI native cloud. A company that was built with only one singular purpose, to provision to host GPUs as the era of accelerated computing showed up, and to host for AI clouds. They've got some fantastic customers, and they're growing incredibly.
One of the platforms that I'm quite excited about is Palantir and Dell. The three of our companies have made it possible to stand up a brand new type of AI platform, the Palantir Ontology Platform, an AI platform. And we could stand up these platforms in any country, in any air-gapped region, completely on prem, completely on site, completely in the field.
AI could be deployed literally everywhere. Without our confidential computing capability, without our ability to build the end-to-end system, as well as offer the entire accelerated computing and AI stack, from data processing, whether it's vectors or structures, all the way to AI, it wouldn't have been possible.
I wanted to show you these examples. This is our special working relationship with the world's cloud service providers. And many Well, all of them are here. And I get the benefit of seeing them during booth tour, and it's just so incredibly exciting. I just want to thank all of you for the hard work.
What NVIDIA has done is this. And you're going to see this theme over and over again. Nvidia is vertically integrated, the world's first vertically integrated but horizontally open company. And the reason that's necessary is very simple. Accelerated computing is not a chip problem. Accelerated computing is not a systems problem.
Accelerated computing has a missing word. We just never say it anymore. Application acceleration. You can If I could make a computer run everything faster, that's called a CPU. But that's run out of steam. The only way for us to accelerate applications going forward and continue to bring tremendous speed up, tremendous cost reduction is through application or domain specific acceleration.
I dropped that phrase in the in the front, and therefore it just became applicate accelerated computing. And that is the reason why Nvidia has to be library after library, domain after domain, vertical after vertical. We are a vertically integrated computing company. There is no other way. We have to understand the applications, we have to understand the domain, we have to understand fundamentally the algorithms, and we have to figure out how to deploy the algorithm in whatever scenario it wants to be deployed whether it's a data center, cloud, on prem, at the edge, on a robotic system.
All of those computing systems are different. And finally, the systems and chips. We are vertically integrated. What makes them incredibly powerful, and the reason why you saw all the slides, is because Nvidia's horizontally open. We'll work and integrate Nvidia's technology into whatever platform you would like us to integrate into.
We offer you the software, we offer you libraries, we integrate with your technology so that we can bring accelerated computing to everybody in the world. Well, this GTC is really a great demonstration of that. You know, most of the time most of the time you'll see me talk about these verticals and I'll use them as examples, but in every single case whether it's automotive, by by the way, financial services, the largest percentage of attendees at this GTC is from the financial services industry.
I know. I I'm hoping it's developers, not traders. Guys, here's here's Here's one thing I wanted to say. And so, in the audience represents Nvidia's ecosystem upstream of our supply chain and downstream of our supply chain. And we work we think about our supply chain upstream and downstream. And it's just so exciting that our entire upstream supply chain this last year, irrespective of whether you're a 50-year-old company, we have 70-year-old companies, we have 150-year-old company who are now part of Nvidia's supply chain and partnering with us either upstream or downstream, and last year you had your record year.
Did you not? Congratulations. We're on to something here. This is the beginning of something very, very big. And so, if you look at accelerated computing, we've now set the computing platform, but in order for us to activate those computing platforms, we need to have domain-specific libraries that solve very important problems in each one of the verticals that we address.
You see us addressing every single one of this. Autonomous vehicles, our reach, our breadth, our impact, incredible. We have a track on that. Financial services I just mentioned, algorithmic trading is going from classical machine learning with human feature engineering called quant the quants did that to now supercomputers studying massive amounts of data, discovering insight, and discovering patterns by itself.
And so, this is going through its deep learning and its transformer moment. Health care is going is going through their chat GPT moment. Some really exciting work that we're there. We have We have a great keynote track here. We have a great keynote track Kimberly Powell is doing a great keynote track um for health care.
We're talking about AI physics or AI biology for drug discovery, AI agents for customer service, and support of diagnosis diagnosis. And of course, physical AI robotic systems. All these different vectors of AI have different platforms that NVIDIA provides. Industrial, we are completely resetting and starting the largest buildout of human history.
And most of the world's industries building AI factories, building chip plants, building computer plants are represented here today. Media and entertainment, gaming of course, real-time AI platform so that we could translation and broadcast support and live live games and live video. Enormous amount of it will be augmented with AI.
We have a We have a platform called Holoscan. Quantum, there are 35 different companies here building with us the next generation of quantum GPU hybrid systems. Uh retail and CPG using Nvidia for supply chain, using creating agentic shopping systems. AI agents for customer support. A lot of work being done here.
$35 trillion industry. Robotics, $50 trillion industry in manufacturing. Nvidia's been working in this area for a decade now, building three computers, the fundamental computers necessary to build robotic systems. We are integrated with working with literally every single company that we know of building robots.
We have 110 robots here at the show. And then telecommunications. About as large as the world's IT industry, about $2 trillion. We see, of course, base stations everywhere. It's one of the world's infrastructures. It was the infrastructure of the last generation of computing. That infrastructure is going to get completely reinvented.
And the reason for that is very simple. That base station, which is it does one thing, which is base station, is going to be an AI infrastructure platform in the future. AI will run at the edge. And so, lots of lots of great great discussion there. And our platform there is called Aerial or AI ran. Big partnership with Nokia, big partnership with T-Mobile, and many others.
At the core of our business, everything that I just mentioned, computing platforms, but very importantly, our CUDA-X libraries. Our CUDA-X libraries is the algorithm the algorithms that Nvidia invents. We are an algorithm company. That's what makes us special. That what That's what makes it possible for me to be able to go into every single one of these industries, imagine the future, and have the world's best computer scientists describe and solve problems, refactor it, re-express it, and turn it into a library.
We have so many. I think we have at this show, we're announcing 100 100 libraries. 70 libraries, maybe 40 models. And that's just at the show. We're updating these all the time. We're updating them all the time. The libraries is the crown jewels of our company. It is what makes it possible for that platform, the computing platform, to be activated in service of solving a problem, making impact.
One of the biggest, one of the most important libraries that we ever created, cuDNN. CUDA deep neural networks. It completely revolutionized artificial intelligence, caused the big bang of modern AI. Let me show you a short video about CUDA-X. 20 years ago, we built CUDA, a single architecture for accelerated computing.
Today, we've reinvented computing. 1,000 CUDA-X libraries help developers make breakthroughs in every field of science and engineering. cuOpt for decision optimization. cuLitho for computational lithography. cuDSS for direct sparse solvers. cuEquivariance for geometry-aware neural networks. Ariel for AI RAM.
Warp for differentiable physics. Parabricks for genomics. At their foundation are algorithms. And they are beautiful. Everything you saw was a simulation. Some of it was principal solvers, fundamental physics solvers. Some of it was AI surrogates, AI physical models, and some of it was physical AI robotics models.
Everything was simulated. Nothing was animated, nothing was articulated, everything was completely simulated. That is what fundamentally NVIDIA does. It is through the connection of understanding of the algorithms with our computing platforms that we're able to open up to unlock these opportunities.
NVIDIA is a vertically integrated computing company with open horizontal integration with the world. So, that's CUDA X. Well, just now you saw a whole bunch of companies. You saw Walmart and, you know, L'Oreal and incredible companies, established companies, JP Morgan and Roche and these are companies in companies that have defined society to today.
Toyota is here. These are some of the largest companies in the world. It is also true that there's a whole bunch of companies you've never heard of. These are companies, we call them AI natives. A whole bunch of small companies. This the list is gigantic. I can't I couldn't This is just a little tiny tiny bit of it.
And um I I couldn't decide whether to show you more or show you less. And so I I made it so that you couldn't see any. And And nobody's feelings are hurt. However, inside this list are a bunch of brand new companies. They're companies like, for example, you might have heard a couple of them, Open AI, Anthropic, but there's a whole bunch of others.
There's a whole bunch of others. And they serve different verticals. Something happened in the last 2 years, particularly this last year. We've been working with the AI natives for a long time. And this last year, it just skyrocketed. I'll explain to you why it happened. This This industry has skyrocketed $150 billion of investment into venture investment into startups, the largest in human history.
This is also the first time that the scale of the investments went from millions of dollars, tens of millions of dollars, to hundreds of millions of dollars, and billions of dollars. And the reason for that is this is the first time in history that every single one of these companies needs compute and lots and lots of it.
They need tokens, lots and lots of it. They're either need They're either going to create and build and create tokens and generate tokens, or they're going to integrate add value to tokens that are available created by Anthropic and Open AI and others. And so this industry is different in so many different ways, but the one thing that is very clear, the impact that they're making, this the incredible value that they're delivering already is quite tangible.
AI natives. All because we reinvented computing. Just like during the PC revolution, a whole bunch of new companies were created. Just as Just as during the internet revolution, a whole bunch of companies were created in mobile cloud, a whole bunch of companies were created. Each one of them had their own standards and we're talking about one of the major standards is that just happened incredibly important.
And this generation, we also have our own large number of very, very special companies. We reinvented computing. It stands to reason there's going to be a whole new crop of really important companies, consequential companies for the future of the world. The The Googles, the Amazons, the Metas, consequential companies that have come as a result of the last computing platform shift.
We are now at the beginning of a new platform shift. But what happened in the last couple of years? Well, we've been watching, as you know, we've been working on deep learning and working on AI. The big bang of modern AI, we were right there at the spot and we've been advancing this field for quite some time.
But why the last 2 years? What happened in the last 2 years? Well, three things. Chat GPT, of course, started the generative AI era. It's able to not just understand, perceive and understand, it's able to also translate and generate, generation of unique content. I showed you the fusion of generative AI with computer graphics and it brought computer graphics to life.
You guys have just Everybody in the world should be using Chat GPT. I know I use it every single morning. Used it to plan this And so Chat GPT was the generative AI era. The second, by the way, generative generative computing versus the way we used to do computing, it's not It's generative AI is a capability of software, but it has profoundly changed how computing is done.
Computing used to be retrieval-based, now it's generative. Keep that thought in mind when I talk about certain things and you'll realize why it is that everything that we do is going to change how computers are architected, how computers are provided, how computers are going to be built out, and what is the meaning of computing altogether.
Generative AI 2023, end of '22, 2023. The next, reasoning AI, 01, which and then took off with 03. Reasoning allowed it to reflect, allows it to think to itself, allowed it to plan, break down break down problems and decompose a problem it couldn't understand into steps or parts that it could understand.
It could ground itself on research. 01 made generative AI trustworthy and grounded on truth. That caused ChatGPT to simply took off, and that was a very, very big moment. The amount of input tokens that was necessary in order to produce and the amount of output tokens it it generated in order to reason, the model was a little bit larger, it you know, of course you could have much larger models.
The model 01 was a little bit larger, not much larger, but its input token usage for context and its output token for thinking increased the amount of computation tremendously. Then came Claude Code, the first agentic model. It was able to read files, code, compile it, test it, evaluate it, go back and iterate on it.
Claude Code has revolutionized software engineering as all of you know. 100% of Nvidia is using a combination of often times all three of them, Claude Code, Codex, and Cursor all over Nvidia. There's not one software engineer today who is not assisted by one or many AI agents helping them code. Cloud code completely revolutionizes the the new inflection.
And the for the first time you don't ask it AI what where when how you ask it create do build you ask it to use tools take your context, read files it's able to agentically break down a problem, reason about it, reflect on it. It's able to solve problems and actually perform tasks. An AI that was able to perceive became an AI that could generate.
An AI that could generate became an AI that could reason. An AI that could reason now became an AI that can actually do work, very productive work. The amount of computation in the last 2 years we know that everybody in this room knows the computing demand for Nvidia GPUs off the charts. Spot pricing is skyrocketing.
You couldn't find a GPU if you tried and yet in the meantime we're shipping GPUs out. Incredible amounts of it and demand just keeps on going up. There's a reason for that. This fundamental inflection finally AI is able to do productive work and therefore the inflection point of inference has arrived.
AI now has to think. In order to think it has to inference. AI now has to do. In order to do it has to inference. AI has to read. In order to do so it has to inference. It has to reason. It has to inference. Every part of AI every time it has to think, it has to reason, it has to do It has to generate tokens.
It has to inference. It's way past training now. It's in the in the field of inference. So, the the inference inflection has arrived. At the time when the amount of tokens, the amount of compute necessary increased by roughly 10,000 times. Now, when I combine these two, the fact that since the last 2 years, the computing demand computing demand of the work has gone up 10,000 times.
And the amount of usage the amount of usage has probably gone up by 100 times. People have heard me say I believe that computing demand has increased by 1 million times in the last 2 years. It is the feeling that we all have. It is the feeling every startup has. It's the feeling that Open AI has. It's the feeling that Anthropic has.
If they could just get more capacity, they could generate more tokens. Their revenues would go up. More people could use it. The more advanced the smarter the AI could become. We are now at that positive flywheel system. We have We have reached that moment. The inflection the inference inflection has arrived.
Last year at this time, I said that where I stood at that moment in time we saw about 500 billion dollars. We saw 500 billion dollars of very high confidence demand and purchase orders for Blackwell and Rubin through 2026. I said that last year. Now, I don't know if you guys feel the same way, but $500 billion is an enormous amount of revenue.
Not one impressed. I know why you're not impressed because all of you had record years. Well, I'm here to tell you that right now where I stand, a few short months after GTCDC, 1 year after last GTC, right here where I stand, I see through 2027, at least $1 trillion. What are we talking about? Now, does it make any sense?
Oh my god. And that's what I'm going to spend the rest of the time talking about. In fact, we are going to be short. I am certain computing demand will be much higher than that. And there's a reason for that. So, the first thing is um we did a lot of work in the last year. Of course, as you know, 2025 was Nvidia's year of inference.
We wanted to make sure that not only were we good at training and post-training, that we were incredibly good at every single phase of AI, so that the investments that were made, investments made in our infrastructure, could scale out for as long as they would like to use it. And the useful life of Nvidia's infrastructure would be long, and therefore the cost would be incredibly low.
The longer you could use it, the lower the cost. There's no question in my mind, Nvidia systems are the lowest cost infrastructure you could get for AI infrastructure in the world. And so the first part was last year was all about AI for inference, and it drove this inflection point. Simultaneously, we were very pleased last year that Anthropic has come to Nvidia.
That MSL Meta SL has chosen Nvidia. And meanwhile, and as a collection, as a group, this represents 1/3 of the world's AI compute. Open-source models. Open-source models have reached near the frontier, and it is literally everywhere. And Nvidia, as you know, today we're the only platform in the world today that runs every single domain of AI across every single one of these AI models in language, in biology, in computer graphics, computer vision, in speech, proteins and chemicals, robotics and otherwise, edge or cloud, any language.
Nvidia's architecture is fungible for all of that, and we're incredible for all of that. That allows us to be the lowest cost, the highest confidence platform. Because when you're building these systems, as I mentioned, a trillion dollars is an enormous amount of infrastructure. You have to have complete confidence that the trillion dollars you're putting down will be you utilized, will be performant, will be incredibly cost-effective, and have useful life for as long as you could see.
That infrastructure investment you could make on Nvidia, you could make with complete confidence. We have now proven that. It is the only infrastructure in the world that you could go anywhere in the world and build with complete confidence. You want to put it in any of the clouds, we're delighted by that.
You want to put it on prem, we're happy about that. You want to put it in any country, anywhere, we're delighted to support you. We are now a computing platform that runs all of AI. Now, our business already starting to show that. 60% of our business is hyperscalers. The top five hyperscalers. However, even within that top five hyperscalers, some of it is internal AI consumption.
The internal AI consumption, really important work like Rexis is moving from recommender systems of tables and collaborative filtering and content filtering, it's moving towards deep learning and large language models. Search moving to deep learning, large language models. Almost all of these different hyperscale workloads are now moving, shifting towards a workload that Nvidia GPUs are incredibly good at.
But on top of that, because we work with every AI lab, because we work with every AI we accelerate AI every AI model, and because we have a large ecosystem of AI natives that we work with, that we can bring to the clouds, that investment, no matter how large, no matter how quick, that compute will be consumed.
And that represents 60% of our business. The other 40% is just everywhere. Regional clouds, sovereign clouds, enterprise, industrial, robotics, edge, big systems, supercomputing systems, small servers, enterprise servers, the number of systems incredible. The diversity of AI is also its resilience. The span of reach of AI is its resilience.
There is no question this is not a one app technology. This is now fundamental. This is absolutely a new computing platform shift. Well, our job is to continue to advance the technology. And one of the most important things that I mentioned last year was last year was our year of inference. We dedicated everything.
We took a giant chance and reinvented while Hopper was at its prime and it was just cooking. We decided that the Hopper architecture the NVLink by eight had to be taken to the next level. We completely re-architected the system, disaggregated the computing system altogether, and created NVLink 72. The way that it's built, the way it's manufactured, the way it's programmed completely changed.
Grace Blackwell NVLink 72 was a giant bet. And it wasn't easy for anybody. And many of my partners here in the room, I want to thank all of you for the hard work that you guys did. Thank you. NVLink 72, NVFP4, not just FP4 precision, FP4 is a whole different type of Tensor Core and computational unit.
We demonstrated now that we can inference NVFP4 without loss of precision but gigantic boost in performance and energy efficiency. We've also been able to use NVFP4 for training. So, NVLink 72, NVFP4, the invention of Dynamo, TensorRT LLM, a whole bunch of new algorithms, we even built a supercomputer to help us optimize kernels and help us optimize our complete stack.
We call it DGX Cloud. We invested billions of dollars of supercomputing capability help us create the kernels, the software that made inference possible. Well, the results all came together and people told people used to tell me, "But Jensen, inference is so easy." Inference is the ultimate hard. Inference is ultimate hard, it is also ultimate important because it drives your revenues.
And so this is the outcome. This is from semi-analysis, this is the largest, most comprehensive sweep of AI that has AI inference that has ever been done. And what you see here on the left on on this side on this side is tokens per watt. Tokens per watt is important because every data center, every single factory, by definition, is power constrained.
A 1-gigawatt factory will never become two. It's physically constrained, the laws of atoms, the laws of physicality. And so that 1 gigawatt of data center, you want to drive the maximum number of tokens, which is the production, the product of that factory. So you want that curve you want to be on top of that curve as as high as you want.
This the x-axis is the interactivity, the speed of inference, the speed of each inference. The faster you can inference the faster you could of course respond, but very importantly, the faster you can inference, the larger the models, the more context you could process, the more tokens you can think through.
This axis is the same as smartness of the AI. And so this is the throughput of the AI, this is the smartness of the AI. Notice the smarter the AI the lower your throughput. Makes sense, you're thinking longer. Okay? And so this axis is the speed, and I'm going to come back to this, this is important.
This is where I torture all of you. But it's too important. Every CEO in the world, you watch, every CEO in the world will study their business from now on in the way I'm about to describe. Because this is your token factory. This is your AI factory. This is your revenues. There's no question about that going forward.
And so this is the throughput. This is the intelligence. Better perf per watt for a given power of data center, the more throughput, the more tokens you could produce. On this side is cost. Notice, Nvidia is the highest performance in the world. Nobody would be surprised by that. They would be surprised by the fact that in one generation, whereas Moore's Law would have given us through transistors 50% two times, Moore's Law would probably give us one and a half times more performance.
You would have expected from Hopper H200 one and a half times higher. Nobody would have expected 35 times higher. I said last year at this time that Nvidia's Grace Blackwell NVLink 72 was 35 times perf per watt. Nobody believed me. And then SemiAnalysis came out and Dylan Patel had a quote. He accused me of sandbagging.
He accused me of sandbagging. He says, "Jensen sandbagged. It's actually 50 times." And he's not wrong. He's not wrong. And so our cost per token Yeah. Our cost per token is the lowest in the world. You can't beat it. I've said before, if you have the wrong architecture, even if it's free, it's not cheap enough.
And the reason for that is because no matter what happens, you still have to build a gigawatt data center. You still have to big up build a gigawatt factory, and that gigawatt factory for 15 years amortized across, that gigawatt factory is about 40 billion dollars. Even when you put nothing on, it's 40 billion dollars in.
You better make for darn sure you put the best computer system on that thing so that you could have the best token cost. Nvidia's token cost is world-class. Basically untouchable at the moment. And the reason that's true is because of extreme co-design. And so, I'm very happy that he named us There was a monkey king token king.
Well, we take we take all of our softwares and as I told you, we vertically integrate, but we horizontally open. We're vertical integration horizontal open. We integrate all of our software and all of our technology, however we could package it up and integrate it into the world's inference service providers.
And these these companies are growing so fast. They're growing so fast. Fireworks, Lin is here, together, they're just growing so incredibly fast. 100 times in the last year. They are token factories. And the effectiveness, the performance, and the token cost production capability for their factories is everything to them.
And this is what happened. This is We updated their software. Same system. And notice their token speeds incredible. The difference before before NVIDIA updated everything in all of our algorithms and software and all the technology that we bring to bear, about 700 tokens per second average went to nearly 5,000, seven times higher.
And so this is the incredible power of extreme co-design. I mentioned earlier the importance of factories. This is the importance of factory. Your data center, it used to be a data center for files. It's now a factory to generate tokens. Your factory is limited no matter what. Everybody is looking for land, power, and shell.
Once you build it, you are power limited. Within that power limited infrastructure, you better make for darn sure that your inference, because you know inference is your workload, and tokens is your new commodity, that compute is your revenues, that you want to make sure that the architecture is as optimized as you can.
In the future, every single CSP, every single computer company, every single cloud company, every single AI company, every single company, period, are going to be thinking about their token factory effectiveness. This is your factory in the future. And the reason why I know that is because everybody in this room is powered by intelligence.
And in the future, that intelligence will be augmented by tokens. So let me show you how we got here. On April 6th, 2016, a decade ago, we introduced DGX-1, the world's first computer designed for deep learning. Eight Pascal GPUs connected with the first generation NVLink, 170 teraflops in one computer, the world's first computer designed for AI researchers.
With Volta, we introduced NVLink switch. 16 GPUs connected with full all-to-all bandwidth operating as one giant GPU. A giant step forward, but model sizes continue to grow. The data center needed to become a single unit of computing. So, Mellanox joined NVIDIA. In 2020, DGX A100 superpod became the first GPU supercomputer combining scale-up and scale-out architecture.
NVLink 3 for scale-up, ConnectX 6 and Quantum InfiniBand for scale-out. Then Hopper, the first GPU with the FP8 Transformer engine that launched the generative AI era. NVLink 4, ConnectX 7, BlueField 3 DPUs, second-generation Quantum InfiniBand. It revolutionized computing. Blackwell redefined AI supercomputing system architecture with NVLink 72.
72 GPUs connected by NVLink spines. 130 terabytes per second of all-to-all bandwidth. Compute trays integrate Blackwell GPUs, Grace CPUs, ConnectX 8, and BlueField 3. Scale-out runs over Spectrum 4 Ethernet. With three scaling laws in full steam, pre-training, post-training, and inference, and now agentic systems, compute demand continues to grow exponentially.
And now, Vera Rubin. Architected for every phase of agentic AI, advancing every pillar of computing including CPU, storage, networking, and security. Vera Rubin and the NVLink 72. 3.6 exaflops of compute. 260 terabytes per second of all-to-all NVLink bandwidth. The engine supercharging the era of agentic AI.
The Vera CPU rack. Designed for orchestration and agentic workloads. The STX rack, AI native storage built with BlueField 4. Scale out with Spectrum X co-packaged optics, increasing energy efficiency and resiliency. And an incredible new addition, the Grok 3 LPX rack. Tightly connected to Vera Rubin, Grok's LPU's massive on-chip SRAM.
A token accelerator to the already incredibly fast Vera Rubin. Together, 35 times more throughput per megawatt. The new Vera Rubin platform. 70 chips. Five rack-scale computers. One revolutionary AI supercomputer for agentic AI. 40 million times more compute. In just 10 years. Now, in the in the good old days, when I would say Hopper, I would hold up a chip.
That's just adorable. This is Vera Rubin. When we think Vera When we When we think Vera Rubin, we think the entire system. Vertically integrated. Completely with software. Extended end-to-end. Optimized as one giant system. The reason why it's designed for agentic systems is very clear because agents of course the most important workload is it's thinking the large language model.
The large language models are going to be you larger and larger and larger. It's going to generate more and more tokens more quickly so it could think more quickly. But it also has to access memory. It's going to pound on memory really hard. KB cache, structured data, QDF, unstructured data, QVS. It's going to be pounding on the mem- on the storage system really really hard which is the reason why we reinvented the storage system.
It is also going to use tools. And unlike humans that are more tolerant to slower computers, AI wants the tools to be as fast as possible. These tools, web browsers, in the future they could also be virtual PCs in the cloud. Those PCs have to be and those computers have to be as fast as possible. We created a brand new CPU.
A brand new CPU that's designed for extremely high single threaded performance, incredibly high data output, incredibly good at data processing, and extreme energy efficiency. It is the only data center CPU in the world that uses LPDDR5, LPDDR5, and incredible single threaded performance and performance per watt that is unrivaled.
And so that's we built that so that it could go along with the rest of these racks for agentic processing. And so here it is. This is the Grace Blackwell uh no, Vera Rubin. Where is it? Here it is. Okay. So this is the Vera Rubin system. Notice since the last time, 100% liquid cooled. All of the cables gone.
What used to take 2 days to install, now takes 2 hours. Incredible. And so, the manufacturing cycle time we're going to dramatically reduce. This is also a supercomputer that is cooled by It's cooled by hot water, 45°. Which takes the pressure off of the data center, takes all of that cost and all of that energy that's used to cool the data center and makes it available for the system.
This is the secret sauce. It is the only We're the only company in the world that has today built the sixth sixth generation scale-up switching system. This is not Ethernet. This is not InfiniBand. This is NVLink. This is the sixth generation NVLink. This is insanely hard to do well. It is insanely hard to do, period.
And I'm just super proud of the team. NVLink, completely liquid cooled. This is the brand new Grok system, and I'll show you a little bit more about it. This system, eight Grok chips. This is the LP30. The world's never seen it. Anything that the world's ever seen is V1. This is third generation. And we're in volume production now.
And I'll show you more about that in just a second. The world's first CPO Spectrum-X switch. This is also in full production. Co-packaged optics. Optics comes directly onto this chip, interfaces directly to silicon, electrons gets translated to photons, and it gets directly directly connected to this chip.
We invented the process technology with TSMC. We're the only one in production with it today. It's called Co-op. It's completely revolutionary. Nvidia is in full production with Spectrum-X. This is the Vera system. Twice the performance per watt of any any CPUs in the world today. It is also in production.
Well, you know, we never we never thought we would be selling CPUs standalone. Um we are selling a lot of CPU standalone. This is already for sure going to be a multi-billion dollar business for us. So, I'm very very pleased with our CPU architects. We designed a revolutionary CPU. And this is the CX9 powered with Vera CPU, the BlueField 4 STX, our new storage platform.
Okay, so these are the four These are the the the racks. And it's connected Each one of these racks the NVLink rack This is I showed you guys this before. It's a super heavy. It seems to get heavier every year. Because I think there's just more cables in there every year. And so So, this is the NVLink rack.
We've also taken this technology because it it is so efficient to create a data center with these cabling systems, structured cables. So, we decided to do that for Ethernet. So, this is Ethernet 256 liquid-cooled nodes in one rack. And it is also connected with these incredible connectors. You guys want to see um Rubin Ultra?
Woo! So, this is the Rubin Ultra compute node. Unlike Ruben that slides in horizontally, Ruben Ultra goes into a whole new rack. It's called Kyber that enables us to connect 144 GPUs in one NVLink domain. And so, the Kyber rack this I could lift it, I'm sure, but I won't. It's quite heavy. This This is one compute node and it slides into the Kyber rack vertically.
This is where it connects into. This is the midplane. The Kyber racks could those four top NVLink connectors slide in and connect into this and this becomes one of the nodes. And each one of these racks is a different compute node. And this is the amazing part. This is the midplane. And the back of the midplane, instead of the cabling system, which has its limits in terms of how far we could drive cables, copper cables, we now have this system to connect 144 GPUs.
This is the new NVLink. This sits also vertically and it connects into the midplanes on the back. Compute in the front, NVLink switches in the back, one giant computer. Okay? So, that is Ruben Ultra. As I mentioned As I mentioned, how about we take this back down? I need the rest of my slides. Oh, it's coming down.
Okay. Thank you, Jenine. This is what happens when you This is what happens when you don't practice. Okay. All right. So, um you saw you Take your time. Just don't get hurt. You saw You saw this slide. You know, only on Nvidia's keynote will you see last year's slide presented again. And the reason for that is I just want to let you know that last year I told you something very, very important.
And it's so important it's worthwhile to tell you again. This is probably the single most important chart for the future of AI factories. And every CEO Every CEO in the world will be tracking it, will be studying it very deeply. It's much, much more complicated than this. It's multi-dimensional. But, you will be studying the throughput and the token speed of your AI factories.
The throughput, token speed at ISO power because that's all the power you have throughput and token speed for your factories forever. And that That analysis is going to lead directly to your revenues. What you do this year will show up precisely next year as your revenues. And this chart is what it's all about.
And then I said, on the vertical axis On the vertical axis, thank you guys. On the vertical axis is throughput. On the horizontal axis is token rate. Today, I'm going to show you this. Because we're able Because we're now able to increase the token speed, and because model sizes are increasing, because the token length the context length, depending on the different grades of uh different application use case, continues to grow from maybe 100,000 tokens input length to maybe millions.
The token input length is growing, and also the output token length is growing. And so, all of these play into ultimately the marketing and the pricing of future tokens. Tokens are the new commodity, and like all commodities, once it reaches an inflection, once it becomes mature or it becomes maturing, it will segment into different parts.
The high throughput, low speed could be used for the free tier. The next tier could be the medium tier. Larger model, maybe. Higher speed, for sure. Larger input context length. That translates to a different price point. You could see from all the different services, this one is free, it's a free tier.
The first tier could be $3 per million tokens. The next tier could be $6 per million
Comments 0