welcome everyone to the information's ti tv my name is akash basricha it is tuesday july 28th there are a couple of stories that i want to flag for you before we get started the information exclusively reported that Kimi K3 maker Moonshot AI is now working on its next model, Kimi K4, and to do that, the company is looking to get its hands on more NVIDIA Blackwell chips despite U.S. export restrictions to China. The information also has exclusive reporting that a state-backed company in China has started manufacturing chip-making equipment that would be competitive to chip equipment giant ASML. Shares of ASML fell after the report was published. Both of those stories are on our website. On today's show, we have the details on how the government's AI framework is nearly finished coming together in the US and what exactly is in it. We'll then discuss Microsoft's answer to Mythos with a new security project. We have a surprising update for you on the Claude versus Codex versus Cursor rivalry. And finally, will NVIDIA have the same rollout challenges with Vera Rubin as it did with Blackwell? We will have the answer to that from our NVIDIA reporter who has some great reporting on that question. It's going to be a great show, so let's get right on into it. My colleague Leo Schwartz, our AI and policy reporter, has some new reporting that the Trump administration is close to finalizing its voluntary framework for AI companies. I want to bring on Leo to share with us what he knows. Leo, welcome back to the show. It's great to have you here. Leo Dionneau, Ph.D.: Good to be here. Okay, so what do we know about the government's voluntary AI framework? Where is it at? Leo Dionneau, Ph.D.: I'm excited to come here. I feel like finally with some good news, it's usually there's no clarity. There's a lot of confusion about what's happening. There could be finally a light at the end of the tunnel for AI companies who are trying to have some clarity around how the government is treating frontier models. So if you remember way back to late May, early June, there was this executive order out of the White House that basically said we're going to create a voluntary framework for companies to share their models with the government before release. The EO, the executive order, also said the government had 60 days to come up with that framework. Somehow those 60 days have already passed. We're nearing that deadline of August 1st, and our reporting shows that the government is close to releasing it. They've been circulating a draft with the top three labs, and we might finally have an understanding of what this process will look like. Okay, and so what's in it? What do we know about that? So that remains an open question. Oh, man, Leah. You come bearing good news, and then we still have more questions. But that is the case with every great news story, is that more questions than answers often. So what we know is that there's two big things that will come out of this framework. The first is, what is the definition of a covered frontier model? So basically, what does a model have to reach in order for it to actually qualify for this program? You might think that's obvious. It's capabilities. It's the size of the model. That's the question that the companies are still trying to hammer out with these key White House agencies, which include the Office of the National Cyber Director. It seems like they're circling around this idea of capabilities. But as we report in our article, there are still concerns from smaller companies that they might be frozen out of this process, that it will only be written with the top three companies in mind. Then the other big question is, OK, once you are a covered frontier model, what does it actually mean? What is this process for having your model vetted by the government? We do have details in the article that it will likely be done with a combination of the National Security Agency and Casey, the small agency under commerce that does model review and testing. But still, what we're going to have to wait for it is that final answer on how it will decide which models are eligible for the program and then how the actual review process works. So big three labs. So we're talking OpenAI, we're talking Anthropic, and then Google. These are the three companies that are working closely with the government on this. Are they largely happy about this? Do they have concerns? Where do they stand? Well, OpenAI and Anthropic obviously both have gripes with the government over how it's treated the rollout of their new models with Anthropic. It was a little more severe with Mythos and Fable. Of course, the government imposed export controls over apparent security vulnerabilities. with OpenAI. The Trump administration ordered them to do a staggered rollout of their latest model, GBT 5.6, which Sam Altman was not pleased about and said as much publicly. So I think just having any sort of actual real policy out there, they're thrilled about. At the same time, we know that they've been pushing for different considerations around these questions about what will constitute a frontier model and how the testing will actually be done. Again, if you're remember back to reporting from way long ago two months ago there was this concern about whether the period of which companies would have to submit their models to the government would be too long and too onerous former ai and cryptozar david sachs famously was able to stage a last minute intervention to prevent trump from signing an earlier version of the of the eo uh so they want to make sure that this will be in place that it will have concrete policy but also that it won't create any unnecessary burdens for them as they're trying to release their new models and win this AI race against China. Is there anything about open source in the voluntary framework as we know it? Yet another question that we're waiting to see. We know and we've reported previously and in this article that open source companies have been pushing for considerations in the executive order or the framework, I should say, both that there could be potential exemptions. Basically, they want to make sure they don't have to necessarily go through this process, which could represent another regulatory burden as they're trying to race ahead with innovation and be able to challenge their Chinese counterparts, which are further ahead right now. But I've also heard concerns that some open source companies want to make sure, as I said before, that the framework isn't written only with GBT 5.6 and Mythos in mind. In other words, they want to make sure that their models can actually go through this vetting process and can be given this essential stamp of approval from the government, which will open it up to all sorts of different provisions laid out in the executive order, like basically being able to work with quote unquote critical infrastructure, whether that's banks or different groups in the country to be able to vet for cybersecurity, safety and other types of programs. So basically, if you do go through the program, there probably will be some advantages that go with it. Great. Well, Leo, I want to thank you for coming on. And that is Leo Schwartz, our AI and politics reporter here at The Information. Microsoft is launching a new cybersecurity-focused project with models that are meant to be competitive with Mythos. We actually reported this was going to happen a couple weeks ago, and it happened. So I want to bring on Aaron Holmes, our Microsoft reporter, to walk us through Microsoft's strategy with this play. Aaron, welcome back to the show. It's great to have you here. Happy to be here. So it's great news, right? I mean, you come on three weeks ago. you tell us project perception, I think it's called. It's going to happen. And here we have the announcement. What is it? What did they end up announcing? Yeah. So basically, this is a new product that Microsoft is pitching as a cheaper alternative to Anthropics Mythos. And that's basically, you know, both of them are tools that companies can use powered by AI models to, you know, test their own environments for vulnerabilities. And the AI can theoretically find them a lot faster than human cybersecurity testers previously could. And specifically, you know, Microsoft is saying this is powered by its own cybersecurity models, as well as a mix of other models from Anthropic and OpenAI. And as a result, it's, you know, cheaper to use and easier for companies to get access to than to Mythos. Whether that ends up being true in practice, we still have to wait and see, but that's the pitch that they're making. So it's supposed to be competitive with Mythos, Does that mean that it had to go through any kind of the same government review process at all? As far as I know, the government hasn't put any of the same export controls that they originally put on Mythos and Fable, partly just because I don't think Microsoft has as much of a controversial place in the Trump administration's outlook right now as Anthropic does. but they have released this broadly or more broadly than Anthropic has meaning you know in theory this is something that will be available to all of Microsoft customers rather than just a select few companies which is the way that Anthropic has been rolling out its mythos and OpenAI has been rolling out its Daybreak model so far But one of the interesting parts of what you talked about in your article today was that they actually are, Microsoft is still relying on OpenAI and Anthropics models for part of this offering, right? Yeah, their approach with this and also with, you know, Copilot and their other AI products is that they will essentially choose which model is best suited to different tasks. So, you know, in some cases, a very capable model from Anthropic might handle writing some code for part of a test, but then it'll hand things off to, you know, one of the Microsoft models or a smaller model for other parts. And as a result, they say that they can drive down the cost and get customers essentially the same results for a lower dollar amount. How is Microsoft cybersecurity business doing, broadly speaking? You have written extensively on this. They have a new leader at the top as well. How big is the business? How fast is this growing? What does this tell us about the strategy overall? So Microsoft cybersecurity business is over $20 billion. That number is actually from 2023, which is the last time that Microsoft told us how big its business is. I know that it's grown since then, but I don't have the exact number. But what I have heard is that Microsoft basically wants to accelerate its cybersecurity revenue growth again. And specifically Hyatt Gallet, who's the new leader of its cybersecurity unit as of earlier this year, has been basically overhauling their strategy and trying to focus more on these AI-powered tools and on software that can help companies protect against newer threats that are also being supercharged by AI. And so we've seen her essentially clean house and kind of overhaul the entire security business at Microsoft. And I think that this product release is one of the first big signs of the direction that she's taking. And so very quickly, Aaron, I know you got to go, but we have earnings this week. What should we be watching for with respect to Microsoft earnings? You know, the big question is just whether Microsoft is going to be able to get a meaningful return from the unprecedented amount that it's spending on AI. You know, the company has already been revising its CapEx guidance for the year up and up and up. And we might see it, you know, go even higher than the $190 billion that they forecast for the year ending in June. At the same time, you know, investors and shareholders really want to see meaningful acceleration in its Azure revenue growth, especially because Google Cloud is growing at a much faster rate than Microsoft's Azure at this point. And, you know, people want to see more sales of Copilot. We know that there's 20 million paid Copilot seats, which is a really small fraction of the 450 million total paying office users that Microsoft has. So if they can't show, you know, kind of all of the above, that they're getting traction and ROI from their AI spending this quarter, it's probably going to put even more pressure on the stock, which has already fallen significantly this year. Grateful, Aaron. I want to thank you for making time for us. That is Aaron Holmes, our Microsoft reporter, here at The Information. The rivalry between Cloud Code, Codex, and Cursor is fierce as can be. For more on who is winning, I want to bring on Laura Bratton, author of our Applied AI newsletter, who uncovered some interesting new data on that story this week. Laura, welcome back to the show. It's great to have you here. Good to be here. Happy Tuesday. Happy Tuesday. So let's talk about the data here that you collected. You had some interesting numbers here. Claude Code, the traction that it's getting, we kind of have talked about Codex and the threat that it offers to that story. What is the reality on the ground? The reality on the ground that I've seen is companies are still sticking with Claude Code in many cases because it's the first tool that was in their enterprise. And some of the data that I found was that among around 200 customers of Anthropics that use Cloud Code, even as cost per user rose because of Anthropics shift to usage-based pricing, hitting companies' bills, they were still continuing to see usage within their organizations grow. And why is that? I mean, we can talk about the magnitude in a second, but the headline, why is, if costs are going up, why wouldn't they just go to Codex or to Cursor? Yeah, I think it's just not that simple. Companies want access to proprietary models and frontier tools from both Anthropic and OpenAI. If they're using those models, they want to see what they can do. And in many cases, companies just adopted CloudCode first, and they're still in the early days of getting engineers to use these tools, and engineers have a preference for CloudCode. So even though, you know, some may be beginning to adopt codecs, they might be experimenting with open source tools, Cloud Code is still dominant in terms of the number of users who are primarily doing their coding work there today. And the data that you collected and got your hands on, I mean, just walk us through some of the magnitude here of how much costs did increase and yet how many more people were still using Cloud Code? Because I think the data itself was pretty telling. Yeah, so the data was from an AI startup, Weave, which provides model routing services. And they showed that, you know, even though among their customers, which in some cases were large enterprises, half were about startups, and then about 15% were medium-sized companies. And it just showed that even as Cloud Code costs rose substantially, usage also increased. And then for Cursor in particular, it showed that even as the median cost per user climbed, usage actually declined 17%. So I think it shows the difference in the pricing power between a tool like Cursor and Cloud Code. Okay, so let's go to the other coding giants that also have decided to start their names with C. We've got the Claude versus Codex versus Cursor rivalry. What's going on with Codex and Cursor according to the data that you found? So in my conversations with enterprises, some said that usage of Codex is actually growing faster than Claude code. In particular, T-Mobile and a database provider, Neo4j, said that their engineers are getting excited about Codex. In some cases, Neo4j told me that they feel they're getting a better return from Codex, which I thought was interesting. This was all Vyves-based. They didn't necessarily have data they provided to me to back this up. But I think it is an interesting point that we're seeing from some engineering teams. So Codex growing fast. And then even though I'm hearing a lot of companies say that they're switching from Cursor to Cloud Code or they previously switched from Cursor to Cloud Code, one of my sources, I found out that Cursor is still seeing a lot of traction. Year-to-date, their revenue from enterprises grew about sevenfold, and adoption of their proprietary coding models is growing. So their Composer and Grok models' adoption or usage of those models grew 11-fold year-to-date. Are those two companies, are they using usage-based pricing? Are their prices going up, or are they staying relatively stagnant right now? So Codex adoption is so new and it is usage-based, but I don't think a shift in, I don't think a shift towards usage-based pricing is hitting those customers in the same way it's hitting customers that adopted one tool and then saw the pricing model change. Cursor changed its pricing model to usage-based in the middle of last year, but that's only just beginning to hit some of the customers I talked to's contracts as they're up for renewal this year. Right. You know, I'm sort of thinking now that Cursor inevitably will, it's looking like it's going to be part of the SpaceX AI ecosystem. If not, I mean, I don't know. Forget the details on whether the deal has officially closed or not. But we pretty much know that Cursor is now part of the SpaceX AI family. I sort of wonder what the strategy will be here with Cursor, because you could imagine a world where Elon says, hey, let's make Kerser the lost leader. Let's get market share quickly. We're burning cash anyway. It doesn really matter We not going to turn a profit for years on this I mean I don know It could be interesting to see how that strategy plays out Yeah I think that absolutely what going to happen We going to see my prediction would be that we going to see Cursor own coding models get better and better as they have greater access to GPUs through the SpaceX acquisition I think it's going to, you know, improve their position a lot. And the Composer models are just going to get better. I think that companies are really eager to see another dominant player in this space other than Anthropic and OpenAI in addition to the open source players. So I think that there's space for Cursor. I don't think that companies are going to want to use Anthropic or OpenAI's models through Cursor. And that's where I think we've seen a lot of the leakage from Cursor customers to Anthropic and OpenAI. is companies that are finding that, you know, if Cursor's going to raise prices, we might as well just go to the quad code or codex, use, you know, a harness, aka quad code or codex that are optimized for anthropic and open AI models rather than going through Cursor. Right. And I guess, you know, the strategy that I just sort of described, I guess that would really have to rely on Grok and that model becoming a lot more advanced and Cursor having at least some kind of its own advantage there, which, as we know from our reporting, Grok itself is facing a little bit of headwinds right now. Yeah, both of those models, I think, will be really important to Cursor's strategy for sure. Right, right. So you mentioned the conversation you had with a T-Mobile executive, and this was kind of interesting because you pointed out in your article that T-Mobile is a company that they are allowing their employees to use all three coding agents. And I read that thinking, is that common or is this unique to T-Mobile? Like what's the reality for most of these businesses? Yeah, it's absolutely a commonplace narrative that I'm hearing from nearly every CIO and CTO that I've talked to. Companies want to give their engineers access to all the frontier tools to open source tools if they can. Even in cases where they're not giving access to every tool, they're at least giving more than one because they really want to see their engineers drive adoption. And I think they want to be able to replace one tool with another if one of the AI labs jacks up prices, shifts costs, if government policies change, and that affects how companies are able to access these models. They want to be able to switch between providers easily. Okay. And this is what happens sometimes. You have a question, you forget the answer. Where does the open source equation fit into all this? Yeah. One thing I noticed is that in my conversations, even just a few weeks ago, if not a couple months ago, what I heard was that companies didn't want to use open source models for coding tasks. There was this assumption that if you use a smaller, cheaper model or an open source model, you're not going to get as great of a result as if you use the latest and greatest Anthropic model. But that has really changed as we've seen companies begin to adopt open source harnesses like KiloCode, OpenCode, PyKline, Hermes, a number of services that basically allow you to switch easily between a proprietary model and an open source model for coding work. And what's becoming clearer is that you don't necessarily need the latest and greatest model to do coding work. It's actually more important to have a harness or a software layer that lets a model act as a coding agent to have that harness basically leverage whatever model is best for coding work, the same way we've seen companies use model routers for other kinds of AI work. Right. And so what I was going to ask you was then going back to the diversity of the models, of the coding agents here with Claude, Code, Codex, and Kerser, do you see this as a winner takes all story in the end? Or is this a story where ultimately it's the enterprises and the customers that will benefit from what inevitably could shake out to be a pricing war in the end. I think more and more we're going to see a tension between the proprietary model providers in Anthropic and OpenAI and open source models that are used for coding work. So we'll see a sort of rivalry between Anthropic, OpenAI, and then these open source harnesses such as OpenCode, KeloCode, Klein. And we're going to see that the harness becomes a lot more important and then the actual model for coding work. The harness, meaning, again, the software around the model itself, basically the application that people are using every day. Yeah, so think of the harness as kind of the arms and legs that use the model, which is the brain, to do coding work. So in some cases, you don't need your brain to be working at 100% capacity to do coding work. You just need your arms to, you know, and your hands to reach in the right directions and pick the right thing with, you know, whatever. Yeah. Although, although with all this whispering that's going on, I mean, now we don't even need arms and legs. You just, you just dictate what you need. And, you know, he's, anyway, I need one of those microphones. Yeah, we're going to come up with a better analogy. We're going to come up with an example. No, no, no, but I think, look, it is certainly, the thing that I can never figure out is what the difference is between a harness and then the application itself. And I guess harness is more of like the middle layer infrastructure between the application and the model. Is that right? Yes, that is my understanding. Yeah, okay. Well, Laura, I want to thank you for coming on and helping us make sense of it. That is Laura Bratton, author of our Applied AI newsletter here at The Information. NVIDIA customers are starting to get their hands on the company's latest Verorubin racks. Our NVIDIA reporter, Phoebe Liu, has some new reporting on the initial reviews those customers have on how difficult they are to install. I want to bring her on to talk about what she found. Phoebe, welcome back to the show. It's great to have you here. Great to be here. Okay, so NVIDIA is rolling out its Verorubin racks to customers. What are the initial reviews of that rollout. How's it going? Yeah, so the big, big asterisk here is that it's still very early to tell, and most of the customers that I've spoken to only have basically test racks, so a handful that they're kind of playing with to make sure they can get them online and running smoothly before getting larger shipments of, I don't know, hundreds and thousands of racks. But right now, I think everyone is cautiously optimistic, or at least almost everyone. Getting the test racks online has been easier for them than at least last time around with the Blackwell Racks, which my colleagues reported on was basically very, very excruciatingly difficult because of how different it was from the previous generation. And what exactly were some of the issues? Remind us, last time this happened, what were the problems that came up? Yeah, so cribbing on my excellent colleagues reporting here, since this was before I joined the information, a few of the biggest issues. One, a big difference between the Blackwell racks and the previous Hopper chips was that it was a full kind of rack scale solution, they like to call it, that connected 72 GPUs and 36 CPUs together, plus storage and networking and all of that. And that's just imagine the cabling that it would take to connect that many chips together. Yeah. And make it talk to each other in like speeds at the order of like milliseconds. That's pretty hard. And I think that was one of the big things where like if one cable failed, it was like, how do we figure out like what's going on? Did it take down the entire rack? Like it's impacting more than just one chip. That was one of the difficult things. Someone I talked to this time around who actually works on some of these large clusters described finding where networking failures occurred, like black magic. I can't imagine how hard that was. Meaning you have this giant data center or maybe this giant sort of row of racks. And again, the racks are literally chips on top of each other. And it's like this whole server is down or this whole row in this data center is down. Where is the fault line? Basically, it's kind of like finding a needle in a haystack. Yeah, like multiple kilometers of cabling per rack, if I remember correctly. So yeah, that's pretty hard. And then also my colleagues reported that there was like an early design flaw that caused delays and then also issues with kind of heat because as racks get more powerful they also get hot because that how power works So the engineering involved in that was also pretty difficult. So is the solution to that, I mean, is that, is there inherently something different about the way that they have architected Vera Rubin? Have they made it easier for customers to install these things, or is it just that the customers are sort of better positioned to install these giant racks in data centers? Yeah, so it's a little bit of both. A couple of the cloud executives I spoke to for this basically said they see this as a third generation product because it's the third kind of full rack product that NVIDIA has released, the first being the first Blackwells, and then the second Blackwells of the GB300s. And now there's Verirubin or VR200, some people are calling them. So because it's kind of the third iteration of this configuration with 72 GPUs and 36 CPUs, they're thinking that this is a little bit easier because they've done something relatively similar before. And then the other thing is that when I was talking to some NVIDIA executives down in Santa Clara a couple weeks ago, I want to say, They were saying that because so much of the Vera Rubin compute tray, so that's one tray with four GPUs on it, is so automated and has fewer cables than in the Blackwell rack, it's easier to get that assembly right on the first try. I think one of the executives said it was like a 95% chance that it would work on the first try versus about 20% for the initial Blackwell trays. Wow. Yeah, I don't know how much that's really been tested at scale, but that was the statistic that he cited. Yeah. I mean, when we talked about these challenges that NVIDIA had rolling out the Blackwell chips a couple, what is it, nearly a year ago now, I think, you know, one of the things that we did talk about with our newsroom and also with guests that came on is that this is like oftentimes the first time that customers have embarked on a data center build out of this scale and installing chips and stuff like that. And so I do think that it's kind of interesting. I think the customers themselves are probably also very happy. It's easier to found a company the second time around. It's easier to build a data center the second time around. And so NVIDIA was sort of paving the path here, which, by the way, we should say, I mean, AMD also has its Helios Rack system that it's announced and it will put in data centers. And so it's sort of like somebody had to be the... I'm the first of three children and somebody has to blaze the trail for everyone else. The AMD Rack is very interesting. I'm curious to see how widely it's going to be adopted. I think people are saying that because that's the first AMD Rack scale solution, it might run into some similar issues that the first NVIDIA Blackwells did. I was at the AMD conference last week and was chatting with people about that. but yeah very interesting um supposedly it's at full production but i don't really know what that means so yeah we'll see so now everything looks like it could be smooth sailing of course as you said it's very early in the rollout so we you know we have to see how this plays out what are some of the caveats here that people are telling you i mean is is is everybody as confident um as some of these folks you talked to about the rollout of the Rubens, or are there sort of some caveats that we should note here? Yeah, I think not everyone is that optimistic, especially as we move from the initial Vera Ruben racks to bigger clusters. I think because networking was one of the issues in the Blackwell rollout, once customers are trying to connect hundreds and thousands and tens of thousands of chips with, I don't know, I guess thousands of racks that comprise them. That will be a bigger test just because it's very different getting one rack to work versus getting 10,000 racks to work. So we'll see as that happens probably late this year, early next year. And then as we move into Rubin Ultra, which is the next generation of Vera Rubin rack, That will probably bring up a lot of additional questions as well, just because it's new architecture. There are some racks that have been discussed that will hold even more than 72 GPUs and connect more together. So there's one rack that would have 144 GPUs and would basically slot them in like a bookshelf, like the ones behind me, rather than like Baker's rack style. the way the existing Verirubon racks and the Blackwell racks exist. And there's another where it's like eight racks with 72 GPUs per rack connected with optical fiber. So that will all produce additional challenges, especially because the bigger racks can draw almost three times as much power also than the early Verirubon racks. So lots of questions, lots of things to find out. And I mean, we should say that the power issue is significant here, right? Because we've had folks on the show talking about you have to find a way to cool the Reuben chips because of the fact that they produce so much heat, they produce so much power. I mean, liquid cooling is the innovation I think that people have landed on. But that too takes a lot of setup, right? You have to get water for one thing. Yeah, for sure. It's very interesting because I think the second generation Blackwell is the first fully liquid-cooled rack. Now networking is also liquid-cooled. And with the Verirubin racks, NVIDIA is saying that it can be cooled with liquid that's as hot as, I don't know, 113 degrees Fahrenheit, which is also new and will be interesting how that plays out at scale. So as you look ahead on this story, Phoebe, what are the questions that you are focused on finding answers to as we follow this rollout? I mean, how will we see the challenges manifest? Is this going to look like data centers being delayed? Is this going to look like, I don't know, like downtime for models? Like what sort of signals are you watching for? Yeah, for sure. I feel like the really big question is, well, there's two things. One, NVIDIA's biggest priority is to make sure that from the day a chip or a rack leaves a production facility to the day it gets turned on and is serving customers, they want to make that timeline as short as possible. and it'll be very interesting to see if kind of issues with supply chain and kind of how fast everyone can move and get things to work would result in chips sitting idle which would be Nvidia's worst nightmare but there's just a lot of engineering required to make sure that doesn't happen so that's definitely something I'm following and then also just timeline like when When are customers getting large scale shipments of these racks? Who's getting them first? How is NVIDIA deciding who gets what? Those are some of the biggest questions I have, I think. Right. And the other point on the supply chain piece is with the memory chip supply chain crunch, whether or not that impacts any of their operations, I think might be interesting. For sure. Very interestingly, NVIDIA has claimed that they got ahead of the memory crunch and that doesn't affect them because they have a team that tracks the stuff and makes sure they have enough and that they started doing this several years out. So they're hoarding them. They have them all in a warehouse. Something like that. I don't know if you all saw their announcement with a $500 billion partnership with the SK Group, which owns memory maker SK Hynix. And one of the stated objectives of that partnership, don't really know how that $500 billion number is broken down, but one of the stated objectives is that it will allow NVIDIA to make sure it can continue to secure enough memory capacity going forward. So that was interesting. Great. Well, Phoebe, I want to thank you for coming on. That is Phoebe Liu, our NVIDIA reporter here at The Information. That does it for today's show. A reminder, we are on this stream Monday through Friday at 10 a.m. Pacific, 1 p.m. Eastern. If you can't make it then, episodes are available on theinformation.com, on our YouTube channel or wherever you get your podcasts. Make sure to follow us on social media, on X, Instagram, TikTok, and LinkedIn. I'm already excited for our next show tomorrow. Have a great rest of your Tuesday. Bye-bye for now.
