Google I/O Goes Full Stack, NVIDIA Prints $81B, and the SaaSpocalypse Debate Reaches Its Verdict | Ep. 305

The handpicked topics for this week are:

1. Google I/O 2026: Gemini Becomes the Operating System. Google I/O repositioned Gemini from a product to the operating layer for everything Google does, and the numbers backed it up. 900 million monthly active users, 3.2 quadrillion tokens per month, a 7x jump year over year. Pat's headline: this is about widening distribution, not just model quality. Gemini 3.5 Flash, Antigravity 2.0, Gemini Spark, and Android XR glasses all extend Gemini into surfaces that no competitor can replicate. Daniel's read: the token-cost reckoning is coming, and when enterprise subsidies end, models that can deliver value at a lower cost per token will become the ground zero of the next era. (The Decode)
2. Dell Technologies World 2026: AI Factory Goes Agentic, 1,000 New AI Server Clients. Pat and Dan were both on the ground in Las Vegas and called it the most consequential Dell event in years. Michael Dell and Jensen Huang co-keynoted to launch the next-generation Dell AI Factory with liquid-cooled PowerEdge XE9780 servers, Dell Deskside Agentic AI, and a multi-model ecosystem including Google Distributed Cloud with Gemini 3.0, on-prem OpenAI Codex, and Grok. 1,000 new AI server clients in a single quarter is the cleanest leading indicator of enterprise demand heading into Dell's Q1 print. Pat's biggest takeaway: OpenShell as a control plane for agents spanning from the GB10 all the way to the PowerEdge rack has been the missing orchestration piece. Daniel's read: large enterprises are going to build hybrid AI architectures and want to deliver tokens at the lowest possible on-prem cost, and Dell is ready. (The Decode)
3. Blackstone and Google Launch a $5B TPU Joint Venture. Pat called it the biggest story of the week and the one that went most under the radar. For the first time, a hyperscaler has released its proprietary AI silicon to a third-party distribution entity. The $5 billion deal, up to $25 billion with leverage, targets 500 megawatts of capacity online by 2027. Daniel's framing: Google decided its custom silicon is worth more as a commercially distributed asset than as a captive moat. Pat's note: the proprietary nature of TPU infrastructure means retrofitting existing data centers will require real work, but the sovereign angle gives the JV a natural first market. (The Decode)
4. AMD Helios, $10B Taiwan Investment, and the MI450 Anchor Customer Rumor. AMD dropped a $10 billion Taiwan ecosystem investment alongside confirmation that Helios rack-scale is on track for multi-gigawatt customer deployments beginning 2H 2026. A Citi rumor surfaced Anthropic as the anchor MI450 customer, to be formally announced at AMD's Advancing AI Day in July. Pat's read: Lisa Su has made a commitment, and she almost never falls through. The analysts who said AMD would not ship anything in the second half of 2026 are going to be very wrong. (The Decode)
5. OpenAI Guaranteed Capacity: Sam Altman's Moment. OpenAI launched multi-year compute commitment contracts the same week that Anthropic was struggling with capacity outages. Pat called it brilliant and said it makes Sam Altman look like a genius. It’s the inference-era analog of cloud reserved instances: guaranteed availability at a locked price for one, two, or three years. Daniel added context: Anthropic's annualized ARR growth is nearly double OpenAI's and is about to lap them, so the model war is far from over. But for enterprises that need reliability, OpenAI just made the most compelling enterprise trust argument of the week. (The Decode)
6. Sovereign AI Crosses $30 Billion at NVIDIA, 14% of Revenue. NVIDIA disclosed sovereign AI as a segment-level line for the first time, at $30 billion in FY26, 3x the prior year. Pat has been tracking sovereign for years and calls this the clearest possible signal that it has moved from marketing term to structural revenue category. Daniel's point: outside of the four or five hyperscalers doing all the major buying, sovereign is where the incremental demand is coming from, and it is very real. (The Decode) 
7. The Flip: Is the SaaSpocalypse Over? Daniel took the affirmative and came in loaded. Every earnings report across CrowdStrike, Cloudflare, ServiceNow, Intuit, Salesforce, Atlassian, Notion, and monday.com shows companies growing with the AI tailwind. His core argument: there was a reason SaaS emerged 20 to 30 years ago. Companies do not want to be in the software business. Vibe-coded flat-file apps with no security, no governance, no data lineage look great in a kitchen demo and fall apart at enterprise scale. The  is over and he is tired of talking about it. Pat's counter: BofA slapped Salesforce with an Underperform at $160, 8% below where it trades. Snowflake is down 35% year-to-date. A senior Dell executive told him Dell will not buy another SaaS system and is tripling internal software creation. The growth question is real, even if the terminal value is not zero. Both agree the tape will tell the real story. (The Flip)
8. NVIDIA Q1 FY27 Results. Record $81.6 billion revenue, up 85% year over year. Data center at $75.2 billion, up 92%. Non-GAAP EPS of $1.87, up 140%. Q2 guide of $91 billion crushed the $86.8 billion consensus by $4 billion at the midpoint. $80 billion buyback authorized, dividend raised 25x. The stock went down after hours for the fifth consecutive time following a massive beat and raise. Pat's read: NVIDIA may be worth $8 to $9 trillion on paper at a sector-average multiple and 75% gross margins held. Daniel's framing: this is the best company in the world, possibly tied with Google, and it is becoming the Apple of this era. He sees a long, safe journey of continued growth vs. speculative dollars chasing quantum and space names that can double in a week. (Bulls and Bears)
9. Intuit: Earnings Beat, Revenue Miss. A 17% workforce cut, raised guidance, and $8 billion buyback were authorized. Pat's emerging thesis: these companies are cutting people to afford tokens. Intuit comes at a moment when OpenAI's ChatGPT finance plugin via Stripe is building an intelligence layer that could sit on top of Intuit's products without displacing them directly, at least not yet. (Bulls and Bears)
10. Lenovo: Record $21.6 billion quarterly revenue, up 27% year over year. The company’s fastest growth in five years. AI-related revenue is up 84% year over year to 38% of total company revenue. ISG returned to full-year operating profit with a $21 billion AI server pipeline. Pat and Dan both read Lenovo's results as NVIDIA tea leaves, a leading indicator of enterprise AI server demand that directly validates what Dell said on stage about 1,000 new AI server clients. (Bulls and Bears)
11. Analog Devices: Record $3.62 billion revenue, up 37% year over year. EPS up 67%. Q3 guide of $3.9 billion crushed consensus by $270 million. Data center up 90%, industrial up 56%, comms up 79%. The $1.5 billion Empower Semiconductor acquisition adds integrated voltage regulator technology that can reduce AI data center power consumption by 10 to 15% while shrinking the power footprint by up to 4x. Daniel's closing point: you can’t build AI servers without players like Analog Devices and Lattice Semiconductor. These essential node companies aren’t boring; they’re foundational. (Bulls and Bears)

Check out all of our Dell Technologies World coverage linked in the show notes, including our sit-downs with Michael Dell, Jeff Clark, and key customers.

Be part of our community. Hit that subscribe button and see you at Computex.

Note to Uploader: Libsyn AUDIO EMBED – added by uploader.

Disclaimer: The Six Five Pod is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded and reference share prices, but nothing discussed should be taken as investment advice. We are not investment advisors.

‍

DANIEL NEWMAN:
Hey, hey, everybody, we are back episode 305 of the six five podcast. What a week it's been. We've been coast to coast to coast once again, a big week at Dell Technologies World, Pat. We were on the ground sitting down with the man himself, Mr. Michael Dell. And then we jetted back just in time to get you in the chair for your big CNBC interview on NVIDIA. We got back just six hours ahead of schedule and you had a chance to get yourself adjusted and the camera and lights all focused in on you. But what's going on?

PATRICK MOORHEAD: 

How are you doing today? Happy Friday. Yeah, Daniel, thanks. Happy Friday to you too. And I appreciate you getting me back. And by the way, a little inside joke. So I'm a color-coded calendar guy, show up like three hours early in the airport and thinking about when we leave Dell Tech World to make sure that I got there. And anyways. Yeah, I did get there. I did get back two hours beforehand, which was enough time for me to prep. No, but it's been a good week, you know, kind of nursing my health back. I've got another sinus surgery coming up on Tuesday, part two, but cranked out, just had an awesome workout. today and HRV is trending upward, which always makes me feel really good. Big week even next week, we've got Computex coming up. We've got just a lot of interesting stuff hitting out there. Heck, even AMD got the got an announcement out before, which we're going to talk about on the show today.

DANIEL NEWMAN: 

Yeah, no, absolutely. A ton happening this week. We had Google I.O. We had the Dell Tech World. We have, of course, AI in the news, NVIDIA earnings. So we'll have a lot to cover here. We'll have a good flip. We, of course, will get into a lot of what happened in the prints this week and bulls and bears. So much going on, and maybe there's even a thinner, more aged and depressed-looking version of me on the coin. We'll see. I've heard a rumor. You know, like the albino. There's a nice light on me now. You know, I'm pretty tan in real life. This lighting on this new camera, though, makes me look really pasty, doesn't it? No, I think it looks good.

PATRICK MOORHEAD: 

I mean, I really, I really do. I mean, at least not blurry. I'm trying to change it. You know, I've got a 4K really high quality camera, but I've noticed on CNBC hits, I'm just blurry. So I don't understand. Must be them. It's not you, it's them. Gotta be. I know. It can't be me. Must be them.

DANIEL NEWMAN: 

Is the bystander thing coming through? Because that's like the McLaren. That's like the human McLaren vein, they call it. Yeah. Shows that you put in the work. I just want to make sure of that. Because that was really the only reason I got a new camera, right?

PATRICK MOORHEAD: 

Right. And what's interesting, we bumped into a team manager of McLaren. Just kind of bumped into him. It was funny, like second time in three months we've bumped into him.

DANIEL NEWMAN: 

Yeah, well, best team in the world.

PATRICK MOORHEAD: 

I mean, after Red Bull, I agree. Is Red Bull even still like a competitor? I don't know. They're, they're crushing it in the 24 hours of Oman.

DANIEL NEWMAN: 

And I heard, I heard there's a chance that they might swap their talent to RB because RB is actually, I think ahead of them in the points right now. I'm not sure. It's pretty close. Oh, I need to call this out though, by the way. Um, because you know, I care a lot about this, but you know, primarily champions arsenal, just want to make sure everyone knows.

PATRICK MOORHEAD: 

Congratulations. Seriously. I'm, I'm really happy for you.

DANIEL NEWMAN: 

There's been a lot of winning this week. That was winning, for sure. That was winning. Yeah, so there's that, there's this, there's so much more, but Pat, we'll talk about where we're at this week when we dive into Dell Tech World, but let's jump into the decode and let's get after it. All right. Google I O, Pat, what happened?

PATRICK MOORHEAD: 

Yeah, so unfortunately, we weren't able to be there on the ground. We were at Dell Tech World, but a lot of news that they came up here. And I think the backdrop is, so first of all, a very successful Google Cloud Next. a month before it. I think the headliner was infrastructure with the TPU V8, two flavors of it, one for training and one for inference. And then kind of roll into Google I.O., you know, sometimes we forget, you know, being on the analyst side that they've got this huge consumer, that's where most of their profit dollars and revenue come from, from ads, from, you know, YouTube and, Big Gmail, but this kind of is a more consumery type of thing. But they did bring out Gemini 3.5. They brought out basically it's Omni, which is a multi-input model. They brought out Anti-Gravity, which is essentially a desktop application, very similar to Cloud Code. I haven't kicked the tires on that yet, but I'm interested in that. They also brought out Gemini Spark, which is essentially and an AI environment that maximizes on workplace, right? You know, they're gonna add some MCP capabilities, but it's really to keep it simple. So Gmail, Calendar, Chat, Action Items, if, you know, not too many people use the video portion of Workspace, but if you do, right, you could pull in notes. I got a little sneak preview of that, but I'm really interested next week when it becomes available for everybody with with an ultra. an ultra subscription. It's interesting, the 3.5 Pro is not available and they pushed that out. And my guess is that it has to align with the new TPUs. But Spark, as you know, from following the model wars, requires less input and output tokens to get it done. And Google talked about saving a billion dollars 1 trillion tokens per day workload to Gemini 3.5 Flash. And they said it saved money. So again, I'll let the consumer folks hit Android XR glasses and stuff like that. But I really think what this is, is widening the distribution. And Daniel, you and I have talked about that a lot. Distribution versus versus quality. The one thing i'm really going to dive into though is it seems a little bit complex right you need to have there are four different plans that are available to get it every one of the things that. That i had i had discussed and it just seems like it's gonna be difficult you know if you get a. an ultra plan or extreme plan or something like that on the three money losing model makers and computer perplexity aggregator, then this just makes it really tough to know what to do. I know why Google is doing it. Quite frankly, it's to align costs with revenue. because they're actually a public company and need to show max profits.

DANIEL NEWMAN: 

Yeah, absolutely. You know, kind of across the list, what stood out to me, just the growth of Gemini, incredible user growth, they kind of get left out a lot of conversations about, you know, it's like cloud and open AI. But Gemini, You've seen it more than double off that 400 million number now, getting very, very quickly close to a billion monthly active users. But Pat, the number that just blew me away was that 3.2 quadrillion tokens per month, a 7x increase. That's crazy. This just goes to show how quickly we are moving from the sort of training model era to the distribution and inference era. It is happening in incredibly rapid pace. I like your point about Flash. I mean, there's a lot of thought now. I think you saw the other day there was a news about Microsoft. I don't know how accurate it was about their canceling cloud code, but like, The fact is that companies that are now really asking employees to go all out and explore token usage are starting to get the bill. I think when you have people just aimlessly working with AI at scale, it becomes very expensive very quickly. And so models that can deliver value at a lower cost per token will become the ground zero for the next era in the battle. That's going to open up opportunities for models like Flash. It's of course going to open up opportunities for open source models and for specialty vertical models. That's going to start to become much more of a topic when everything's been subsidized and underwritten for so long. Of course, you and I are going to use cloud Opus 47 all day long if they're like, hey, you have all the tokens you want for 200 bucks a month. But when all of a sudden that bill comes in at 20,000 or 2 million or whatever your enterprise scales up to, I think that's going to be a big thing.

PATRICK MOORHEAD: 

Um, yeah, isn't that funny? We kind of went to, we're trying to convince people to use AI to, um, I think it was Uber that said they use 25% of their tokens. And I, I'm sorry if you mentioned that already, but it is, it is, it is pretty, uh, pretty ironic. But then again, um, Anthropica is so off base with how much they bet on infrastructure. I mean, they must have had no confidence on Opus 4.6. because they're off by a factor of probably 10 on what they need to serve that effectively to people. Thank goodness no one's using Grok.

DANIEL NEWMAN: 

Yeah. So funny. Yeah. That's all right. But anyway, so that was my big read. Google continues to have a big advantage with its distribution and its use. People are, I think, Underestimating it consistently. I mean, I mean, you know, now it is evaluated like getting close to 5 trillion. So people aren't necessarily underestimating Google. I just, I feel like they're not really being brought into the conversation, but there is a strong argument that they are the most. Important company and i right now there's an argument to be made just by their share scale and the shares top bottom infrastructure stack that it has. I'm in the of course use the consumption model in the cash flow create so what will come back to you will talk more about google drop the show but. Let's bounce over to Dell Technologies World, Pat. For everyone out there, there's a ton of 6.5 coverage. So Pat and I were at the event on the ground, check out the interview we did with Michael Dell. We sat with Jeff Clark. We sat with a number of the other executives, some of the customers, great conversations. I think you and I probably four or five times highlighted this, but for the people that didn't listen to that, I don't know why, I don't know what you were doing. What was red hot to me? I mean, look, this was all about the pivot with the Dell AI, NVIDIA Dell AI Factory, tokens at scale, ROI on AI, and bringing customers to the forefront. Had some really interesting customers like Eli Lilly came on stage, and Honeywell came on stage, and Samsung came on stage talking about the depths of the partnership. But I mean, what stood out to me, I'd say a couple of things. One is the 1,000 new AI server clients in a single quarter. What a tremendous growth number for the AI factory business. The revenues and backlogs are just incredible. Insatiable demand in this company. They kind of led the way in this whole AI factory narrative. Now you see all the OEMs have kind of followed suit with their own version of the story. The other thing is, I would say, the bringing AI on-prem. Going back to what I just talked about with Google, the thing that really stood out to me, they, of course, had the distribution deal with Google to bring Google on-prem. That was about a year ago. That kind of hit. But now they're talking about with Gemini, with OpenAI, with Rock. I wouldn't be surprised if Anthropic follows, Palantir, ServiceNow, basically bringing the iFactory on-prem to bring token economics into a, you know, going back to what we said about Google, to a price point where you pay for the infrastructure, you deliver your tokens at whatever the cost is to run your infrastructure. You know, the premium is wiped out. And I think that's kind of the underlying story right now as we pivot from Is AI important and useful to, holy crap, it's super useful, and we want to do everything we can with it to scale our productivity and business? How do we do this economically and soundly? And I think large enterprises are going to have a very hybrid architecture. I think they're going to bring a lot of AI infrastructure on-prem, and they're going to want to deliver tokens at the lowest possible cost. And it seems that Dell's on the front foot in terms of offering this to the market.

PATRICK MOORHEAD: 

Yeah. I mean, you nail, I mean, my single biggest takeaway and listen, I don't know if we do a victory lap on this one, but I've been saying for a decade, you know, the ultimate, uh, end point, uh, for enterprise infrastructure is going to be hybrid, right. It's going to be a combination of, uh, on-prem, uh, Colo, Neo cloud, and, uh, uh um is services from from hyperscalers and sass and that's just you know that's just the way it is albeit the growth has been on the hyper biggest growth has been on the hyperscaler side and not on the on-prem enterprise side, I'll say hyperscaler and NeoCloud side. I do believe that even though Anthropic didn't show up, I've got a pretty good idea or feeling, I'll put a 99% chance that Anthropic will be at the table. Now, details were sparse on exactly what OpenAI and Dell were going to do and when. But the fact that the leading model makers are going on-prem and Dell is the infrastructure for that, I think is a huge deal not only for the market, but also for Dell. And the other thing is, as we haven't talked about yet, is doing development on a client, right? And I think it's clear that Nvidia and Dell have a special relationship. I feel like Nvidia views Dell as their primary enterprise go-to-market partner. And given all the distribution that Dell has on the data center side and the PC side, but we saw. And this is this is key here and it really didn't go ding ding ding for me at GTC and that's open shell until I got to Dell so. Open shell is a control plane for agents. and it allows the enterprise to determine agents that go from all the way from this GB10 to the GB300 and workstations, all the way to the power rack with PowerEdge. And you want to do tokens on-prem, but let's say you run out and then you go to the public cloud or you go to the Neo clouds. And that's been a missing piece, essentially a model orchestrator. The next thing above that we need is an enterprise agent orchestrator to fully align through that. Technologically, those two don't have to be from the same company, but there does need to be that layer. And there's a lot of smaller companies who are doing this, companies like Cohere for the enterprise. the ability to arbitrate based on where you want to do those tokens, which would be based on capability, which would be based on availability, which would be based on security, where for some agents, you might never want to put that data up into a hyperscaler model. But all in all, it's a really good show. That is for sure.

DANIEL NEWMAN: 

Yeah, absolutely. Great show. Everybody check out the coverage we had. We had a ton of good content there. We'll make sure in the show notes that shows up so you can see the different interviews. Pat and I worked like dogs. We were running around there.

PATRICK MOORHEAD: 

Yeah, we did eight. I mean, I think we had maybe three hours of downtime the entire show between, you know, and getting up early and late dinners.

DANIEL NEWMAN: 

Yeah, we did the eight videos there. Nine, if you count. Did they have you do the little stand-up thing?

PATRICK MOORHEAD: 

No, no, they chose you for that. You're the king. They had Ben do it, didn't you? How'd you get on with that one? No, no.

DANIEL NEWMAN: 

I'm going to complain and choose me.

PATRICK MOORHEAD: 

It's torture. I'm not saying anything. They want to sign you up for more. I'm game. Puke.

DANIEL NEWMAN: 

No, I'm kidding. That was a ton of fun. I did seven TV interviews in the last 24 hours, too. So I'm just amazing.

PATRICK MOORHEAD: 

You know, I turned, you know, I got 11 p.m. offering. I got like four o'clock in the morning offering and I just. My HRV is sucking, which it is right now. It's literally a lifetime low, at least since I've been tracking. I'm just going to say no. You got to say no.

DANIEL NEWMAN: 

Yeah. The worst one I did was 920. I didn't have any crazy 920 at night. And the earliest one I did was 7 AM. So I didn't have any terrible, because I've done that. I've definitely been that guy. You know, I love a little attention when I can get it. But, you know, 4 a.m., I call it the night show for the West Coast. It's the evening tech show for the West Coast. Anyway, so let's talk about another big piece of news, maybe one of the most important structural stories of the week, which was Blackstone and Google partner on a JV on round TPUs. What is that?

PATRICK MOORHEAD: 

Yeah, it's so funny. This kind of went under the radar. And I think it was probably the biggest story of the week. Maybe we should have led with this one. But there's been a lot of talk about, does TPU go outside of Google and how? We saw Anthropic did a deal with Google. But that is actually in Google data centers. And the thought was, will it ever break out of a Google data center? And these are great questions because the TPU is really good. First of all, has lower performance per chip than NVIDIA and AMD. But the magic here is it's scaling. The network capabilities that it has are incredible. So the aggregate performance of what you can do is pretty awesome. But it's massively proprietary right whether it's water cooling all you know it is absent the racks are proprietary. There's nothing so you can't just roll this thing into you know a standard i don't know core we've core we've data center they're gonna have to make. some changes inside an already built facility. But this Blackstone and Google, it answers the question, right? $5 billion deal, but up to $25 billion with leverage, a.k.a. debt. It's a joint venture, makes a lot of sense. And what a twist on creating your own NeoCloud, right? And this clearly says that Google wants to actually make a real business about renting TPUs. And that's a good sign in that they think they can actually make money off of this outside of reducing their costs like they're doing internally for Google SaaS properties like YouTube and Gmail and things like that. Also, final comment kind of layers into the, layers into the whole sovereign element, right? So 500 megawatts to start 2027, turning this thing alive, that's actually enough time to do a kind of a bespoke as opposed to renting room or retrofitting something from one of the Neo clouds, like a nebbiest or a core weave.

DANIEL NEWMAN: 

Yeah, no, absolutely. I think you hit it pretty well. Not a lot for me to add here. I think it was an interesting decision. I don't think anyone else has done this. I mean, I could see AWS going down this path, but the hyperscale decided it's custom silicone is worth more as a distributed asset than as a captive moat. That's an interesting question. interesting to see how big they want to how much they want to focus on the custom silicon business and now it's no longer a homegrown for our own purposes but it's becoming a distributed product and doesn't you know that whole narrative that's run along about google competing with nvidia get more legs here you know because as long as it was sort of a pure in-house play to me it was always like of course they're going to do that but now that they're going to go out is google basically saying we think the market's a lot bigger um and you know but then if you have to weigh in how much capacity it has, how it best applies capacity. Google makes money by putting its infrastructure to work for itself. That's the number one place it makes money. The next place it makes money is offering it as part of its cloud services to the customers in the cloud. And now basically, you know, starting a new venture, bringing it outside is, you know, how big does this business get? But I think it all goes to the, you know, the three plus trillion of expected infrastructure in 2030. They want to maybe make a little more picks and shovels. It's another ROI element, another way to drive cash for the business, and another way to deepen its criticality in the full AI supply chain. Google wants to control its fate. It just keeps vertically integrating further and further and further up and down the stack. So I can't fight it. This company made $57 billion just on its own investments last quarter. The heck do we know? All right, let's bounce on to something. And I'm going to be a little brief on this one, because I haven't just candidly been on the road, have not had a lot of time to read up on it. I'm sure you have. So AMD, Helios, it appears that this is rapidly moving forward, Pat. There was a new $10 billion investment into Taiwan. Lisa's out there securing more capacity. Um, you know, we're, we're seeing, uh, next generation Epic running on TSMC two nanometer. Um, so we've got the architecture now, Helios semi four 50. You've got, they've got ecosystem investments going on here. Um, and I believe they named an anchor tenant and drop it. For the MI four.

PATRICK MOORHEAD: 

Yeah, that was a, that was a city bank, a city bank rumor.

DANIEL NEWMAN: 

Okay, so let's be clear. I'm spreading FUD, maybe. No, not FUD. I'm spreading rumors. So that may or may not be true. I had it in my notes that it could be true. Glad I asked. But, you know, it appears that, gosh, AMD, Pat, and by the way, our equities team initiated the highest street price target on AMD that exists right now. I think almost $800 a share is what the team called. I don't know. So the market clearly thinks AMD is going somewhere and it looks like they are. Give a little more on what happened though. I just haven't had the full chance to. Yeah.

PATRICK MOORHEAD: 

Well, here's the thing. There were three things hit at the same time, right? A $10 billion investment into Taiwan, that's for 2.5 de-packaging for higher interconnect bandwidth for Epic Venice. Then the second was Helios Rack. They confirmed it's on track for multi-gigawatt customer deployments beginning 2H26. They listed the ODM partners. Then finally, this Citi rumor that says that they'll announce Anthropic at advancing AI day. By the way, I love that Lisa is getting ahead of the news at Taipei, kind of got their own moment. I don't think that's going to be it. One other thing that AMD launched was a new version of their AI SoC that has the UMA AI400, I believe. And it's essentially a follow on the The 300, they also are going to start selling their own developer box, sided up against the GB10, essentially. So I can't wait to get my hands on this. Listen, I know a lot of analysts said there was no way that AMD was going to ship anything in the second half of 2026. Very reputable firm, and that's ending up to be very, very wrong. So, and leave it up to Lisa to make a commitment and, and follow through very rarely. Does anything Lisa says, falls, falls through the cracks or, or it doesn't happen. It's kind of her, her own brand name.

DANIEL NEWMAN: 

Yeah. AMD, there's a lot to be excited about there. The market is exploding in size. They've got an absolute tailwind with CPU demand with as CPUs and GPUs goes near and near to one-to-one. The market didn't even necessarily realize what a good position it was going to be in to be able to capitalize on this next generation of rack scale. I mean, we'll talk about NVIDIA's $20 billion view into CPU potential business for Vera, but Lisa's CPU business wasn't even in most of these calculus of all the people that thought they were going to be market on the GPU.

PATRICK MOORHEAD: 

The stock is rocking again, man. What is AMD at like 414 or something like that? It's rocking.

DANIEL NEWMAN: 

Yeah. I'll have to see what Rolf. I think Rolf has him at like 780.

PATRICK MOORHEAD: 

I'm sorry, 469. Up another four points today. By the way, Dell Tech is up 15, 16 points today. Crazy.

DANIEL NEWMAN: 

Yeah. Jeez. I think that was something to do with us.

PATRICK MOORHEAD: 

Yeah, it was. I mean, we clearly pointed it out there with the 84% cost savings on-prem versus API that one of our joint babies, Signal 65 testing company, cranked out.

DANIEL NEWMAN: 

Crank it out, crank it out, crank it out. All right, Pat. So open AI is giving the middle finger to Anthropic as Anthropic continues to struggle with its capacity. And Sam saying, hey, everyone questioned my spending and my crazy deals and all the capacity I went out and bought, but right now I'm able to deliver. And you're not, right? I mean, that kind of what he's saying, open AI guaranteed capacity.

PATRICK MOORHEAD: 

Yeah, I mean, I love this, you know, opening up. Well, first of all, I think Codex rules over Claude Code. I mean, everybody I've talked to, I've seen people pivot off of it. NVIDIA is doing a ton with that. And my own son has pivoted from, he went from Cursor to Claude Code to Codex. I see a lot of the active analyst firms in this pivoting over. And it but but for a lot of other workloads it does good at. Numbers financial analyst but i'm stuff like research doesn't do as well but with that said you have to have the capacity even if your model isn't the best. If you're having outages like Anthropic, if you are nerfing your models like Anthropic admitted to, and then to come out and say guaranteed capacity, I mean, I think that's absolutely awesome. It just makes Sam Altman look like a freaking genius. And we were all wondering, why do we need all this capacity? except for you, of course. But this is guaranteed capacity. It's very similar to what we call a cloud-reserved instance, where you go in and if you make a three-year commitment, you get an availability guarantee and you get a price guarantee, typically lower. I don't know if you have to lower the price on this, but getting guaranteed availability Is is important and maybe companies like uber you know wouldn't have bailed so hard had they not done this so yeah i think it's kind of a win win right price and power capacity rationing and. Gosh good luck anthropic and quite frankly even google repeating this already tell the way that. Google has segmented so much of their services and pricing. They're not out of the woods yet. I don't think TPU V8 is fully at capacity yet, but who knows? We might see that. It's a brilliant move. It really is.

DANIEL NEWMAN: 

Yeah, I mean, OpenAI needs a brilliant move. It's business and growth. I think I saw something like Anthropic growth rates almost double. It's about to lap OpenAI in terms of its annualized ARR. Now, again, they do a little different calculation there. So you've got to put context around that. But it ain't over yet. I just want to be clear about that. I, I enjoy a little bit, you know, slapping Sam around. I think he's kind of smug sometimes, but, uh, dark, dark. It was not much better with his fear mongering nonsense. Um, kind of want Google to win. I don't know. I mean, at least they, they don't do anything evil, at least. All right. Isn't that their whole thing? Don't be evil.

PATRICK MOORHEAD: 

Yeah. I mean, at least that's what it says on the poster.

DANIEL NEWMAN: 

Yeah. So don't be evil. I might be evil later today, thinking about it. All right, so let's go to topic six. This is something we'll kind of call out a little bit in our flip. We're going to move a little quicker through these next couple of things. We've still got a lot of ground, and you and I, we've been yapping a while here. you know get some get some measurement and videos earning so what sort of hit this a little bit separately but just a quick touch. You know with enterprise sovereign i am like the whole break out thing that's going on like everyone's kind of attributed all the it's a hyper scalers and we're starting to see some numbers that suggest there's a lot of a build happening i mean the numbers the enterprise numbers. the whole NeoCloud numbers coming out of NVIDIA and a $30 billion Sovereign AI number came out this quarter, 14% of their mix, 3x a prior year. It looks like it's moved from a talking point uh, for Nvidia to a segment line disclosure or, you know, it's getting where it's at least baked into that segment line. Um, I thought it was pretty big deal, Pat. I think, uh, where does revenue come from outside those? Cause it's not even the max seven. It's really like four, four companies, right? Five companies doing all that buying. So sovereign is a big deal. It's very real.

PATRICK MOORHEAD: 

Yeah, it is. It was good to see that. Like it, it, there's no way that there's no other better way to signal that something is real, that you actually call it out on an earnings call. Right. Which is exactly what, uh, what Nvidia did. And, you know, I've been tracking sovereign for, for a long time. Um, and with the new administration coming in and you and I had a lot of conversations. The last two Davos is about sovereign clouds and you know these countries don't wanna get cut off and also they want the data protected and in its own sandbox that a company can't go in. and either turn off the functionality or look at the data, right? And there's three or four different ways to define sovereign. I took a week long trip out to the UAE and I met with core 42 that is right now likely the largest institution out there. They just cranked out They've got a five gigawatt campus that's scheduled, and they're about to go live this summer with a 200 megawatt institution, and they just got their new BlackBull GPUs in. They made a big deal over that. Also Humane over in Saudi Arabia, they're a little bit slower along the way, but Microsoft is absolutely on a tear here. I'm being the layer of sovereign software layer which the country's put another layer on top to make sure it's truly sovereign. But also final point is the definition of sovereign has expanded a bit to even if you're inside the united states. and you're a healthcare company, I see a lot of people standing these up in a very similar way that they did with high-performance computing installations. So yeah, I don't know if this is a hardcore victory lap, but it's real. It's not a marketing term. And quite frankly, five years, it was marketing, okay? I think Oracle really drove the way on this. It's interesting. They're not as much in the conversation as they used to be.

SPEAKER_00: 

Yeah, no, I think that's absolutely right, Mr. Moorhead. We can skip number seven if you want.

DANIEL NEWMAN: 

Yeah, last just quick thing. Anthropic raised a crap load of money at a massive valuation, $50 billion, $1.2 trillion. And they declared that they now have a 37% enterprise share ahead of OpenAI. So nobody going into this year would have thought that there'd be $2 trillion AI labs going public. OpenAI is filing for its IPO, I understand, this week. Anthropic should not be far behind and do $1.2 trillion post money raise. Crazy, absolutely crazy. I remember 850 on the open AI just a few weeks ago. I don't think anyone would have thought.

PATRICK MOORHEAD: 

Yeah, great call out on the market share piece. It is mind boggling on that. Now, those are Dario's numbers, not industry numbers that that I've seen. And my instinct tells me that, you know, the momentum has shifted against them. For the time being given how good that codex is and i don't know if it's because the algorithm but x just keeps telling me i'm leaving cloud code i'm going to i'm going to i'm going to i'm going to codex and. I think enterprises really need to think about trust on directly connecting with Anthropic. I wrote a Forbes article a couple of weeks ago on this. Check it out if you're interested in my view on trust in Anthropic for the enterprise.

SPEAKER_00: 

Absolutely. Absolutely. Where are your token dollars going these days? They're spread.

PATRICK MOORHEAD: 

I'm doing more on opening i am same on on computer and i found that even how crappy the experiences on opus for it does the best research. It just does. Oh, crap.

DANIEL NEWMAN: 

You got to keep giving that shitty company money. I know, dude. It sucks. I hate it. We give them most of our token dollars are going to drop it right now.

PATRICK MOORHEAD: 

So I thought a sophisticated shop like yours would be shifting over to Codex a little bit more.

DANIEL NEWMAN: 

Nope. Nope. My team is locked in as a VMware certified infrastructure implementer, is still committed to their whatever you call it, their brand. I am a Salesforce approved developer. I am a tribal man.

PATRICK MOORHEAD: 

I know how much you love Salesforce too. I do. By the way, everybody, we've got more coming up in June.

DANIEL NEWMAN: 

Yeah. So, you know, we are a Salesforce shop here. I put a lot of dollars in their direction. Um, all right, buddy, time for the flip time to get destroyed. Time to be owned.

PATRICK MOORHEAD: 

Somebody's getting down. You're going down.

DANIEL NEWMAN: 

He's getting grown today. Um, And here's the topic. Is the SaaSpocalypse over? Or is B of A, split call, ServiceNow buy, Salesforce underperform at 160, the real signal that platform-analyzed personalization bifurcates rather than lifts everyone? All right, Pat. So SaaSpocalypse is over or not over? And it's going to continue on. Let's see who's saying it's effectively over. Man, this one's hard because everybody knows where I stand on this. Of course it's not over. I mean, it's over. Yes, yes, of course. The SaaSpocalypse, as far as I'm concerned, is over. It was a rewriting and it needed to be done because many SaaS companies and the whole license-based, seat-based approach doesn't work in an era where you have agents and thousands of agents. So we had to have a moment where we could inflect to see whether or not companies can make that pivot. Look, the earnings are out. It doesn't matter if it's the security companies like CrowdStrike and CloudFlare, the CRMs, ERPs, and workflow companies like ServiceNow, Intuit, Salesforce. Everybody's hitting their numbers. And so I don't care what you say. You can say, oh, I vibe coded my new application. Look, Atlassian, Notion, freaking monday.com, the ones that even seem like you might be able to vibe code them. All of these companies are growing with the AI tailwind because one very simple reason Companies don't want to be in the business of building software. There was a reason 30 years ago when the SAS revolution and the enterprise software revolution began. Was it 30, 20? I don't know. It's been a long time. What happened is companies have been so busy trying to develop their own software systems that it became an overwhelming job for them and they became IT shops and software shops. And guess what happened in enter CRM, enter ERPs, enter HCMs and all these other tools. Because what it allowed businesses to do was focus on doing their business, not focus on the development and building of software. And a lot of people aren't thinking about this because while they're busy vibe coding a cool flat file software that has no real connection to any databases, no security, no sovereignty, no governance, no data lineage. It looks really cool when you demonstrate it to your kids in your kitchen at home. But what happens is every single day businesses evolve and you need big staffs and teams of people to continue to keep code modern in order to keep your data secure, in order to keep transactions flowing. People are quickly realizing that They still need their software providers to do all this. Sure, in the future, could things become headless? Could monetization strategies change? Could we be charging on tokens and outcomes and actions instead of based on seats? Of course. But the SaaSpocalypse is over because enterprises aren't going to be entering the business of actually building software anytime soon. So what's going to end up happening is, of course, like any industry, as it matures, the best companies will continue to grow, evolve, and get stronger. They're going to use the AI to actually give themselves further advantages, build better moats, deeper moats, and become more critical to businesses. And sure, some crappy companies that built crappy products in a crappy era of crappy software might eventually dissipate. But that doesn't mean that they were a victim of the SaaSpocalypse. That just meant they didn't add value in the era of taking SaaS, adding AI, doing it on a token action-based and making it consumable the way the cloud is. So the rewriting is real, the SaaSpocalypse is over, and I'm glad because I'm really fucking tired of talking about it.

PATRICK MOORHEAD: 

So Daniel, I'm just glad.

DANIEL NEWMAN: 

Beat that up, editors.

PATRICK MOORHEAD: 

Dan, I'm just glad to know that you're long in your investment portfolio on these SaaS guys. You've got to pump these guys up. So for the simulated argument, I mean, it's super clear. So first of all, SaaSpocalypse, we've only seen the beginning. There was a little bit of a respite, and we do see some acceleration. But quite frankly, not a large boost on the AI front. Maybe a couple of companies. I think Salesforce. is moving the needle on this. But Salesforce, they just aren't believers in this. Bank of America's price target is 8% below where the stock trades today. Snowflake is down 35% year to date. OpenAI is coming on with this guaranteed capacity and we just talked about Dell and all the on-prem capability to kind of move in a different direction and invest dollars into tokens to build your own software. Heck, Dell Tech World, a very large, a very important senior executive at Dell, told me that they are tripling their amount of internal software creation. And that, you know, I won't cite the exact M in the SAS, but they're going to be. Essentially, they're not going to buy another one of these systems again. I don't think that the SaaS companies, like the top ones, are going to go out of business. I think the only thing that we need to question is the growth. And if you look at how unintegrated AI is in the enterprise today and you project forward, I think we're kind of at the, the base of the iceberg and nobody believes it at this point at least. at least in the market. So the other side of my mouth though, right now, there have been some pretty good revenue numbers that come out and profit numbers. I am expecting the SaaS companies to do mass layoffs and even over and above what Intuit has done. And potentially that will get the companies in a position where they spend a whole lot less and can do a whole lot more. I'm also expecting massive reductions in marketing. If you look at SaaS, marketing as a percentage of revenue, it's higher than almost any other business that's out there, except for consumer packaged goods. So yeah, I think we've just seen the beginning here. You can't hedge.

DANIEL NEWMAN: 

You're not allowed to hedge. Okay, no hedging in your answers, but that was some good hedging, nevertheless. All right, we got a ton to cover in Bulls and Bears, Pat. I think I won that one.

SPEAKER_00: 

I'm just kidding. I don't know.

DANIEL NEWMAN: 

You know what's going to actually tell the truth is the tape. The tape will tell what ends up happening here because… I don't know if it's the tape. Yeah.

PATRICK MOORHEAD: 

Because it hasn't actually… Sorry, go ahead. No, I'm going to be the stock price.

DANIEL NEWMAN: 

It's going to be the print, right? I mean, yeah, I meant that as in the tape of the yes, but you could argue both. But the voting and weighing machine is real. And I mean, the market is going to say what it said in the rating is real. Like these companies will never trade at the valuations that they did unless the market truly believes that they're pure native. And even then, Google and these companies trade on a multiple of EBITDA. They don't trade on free cashflow and revenue. That's not how those companies… Let me ask you this.

PATRICK MOORHEAD: 

Not that it's associated with you at all, a company called Futurum Equities. Is Futurum Equities doing strong buys on SaaS companies right now?

DANIEL NEWMAN: 

Well, our initiation and coverage is all focused on semis and infrastructure right now, but when we announced and launched the coverage on SaaS companies at some point, I will have to ask those what they say, because I don't touch that part of the business. I just parent that part of the business. But we do have a strong buy on AMD.

PATRICK MOORHEAD: 

There we go, baby. Not you guys. Futurum equities. That's not directly affiliated with you. Looks like geniuses.

DANIEL NEWMAN: 

Yeah. Yeah. I mean, so right now, I mean, why would you initiate coverage on that when you could initiate coverage on infracompanies? Everybody loves infracompanies. All right. So let's speaking of initiating coverage and stocks and markets, let's go to bulls and bears. want to change it to bulls and bubble bears. Bulls and bubble bears. Bulls and bubble bears with a rocket ship. Rocket ship, rocket ship, rocket ship. Bite me, bubble bears. All right, Pat, you covered it so well on the telly, telly, telly. By the way, that guy, it was really funny when the guy you were paired with started reading the tape. Oh, man. Like, like, like, they were like, all right, guy. We'll go ahead and do that, you know. Anyway, what happened with NVIDIA?

PATRICK MOORHEAD: 

Yeah, so first of all, revenue beat, non-GAAP, EPS beat, data center beat, I mean guide beat, and then massive re-ratings to the positive side. I don't think anybody brought anything down. Typical price was around 300 bucks. And then the stock did exactly what you would expect based on, it went down. Actually, this is the fifth out of five out of five. Last time they've had great beats and guidance that the stock has gone down. So nobody should be surprised here. I was looking at a couple of things here, right? I was looking at China, I was looking at gross margins, and we did get some answers on that, right? 75% gross margin held, right, which was good. NVIDIA has zeroed out China. So in other words, they didn't have to go back and do some massive write-off again or something like that, but had a massive guy, $91 billion. Think about that. Are they going to be a $500 billion run rate company revenue-wise. That's absolutely mind-blowing. One of the bigger things that Nvidia did, did an $80 billion buyback and they 25Xed the dividend. Reminding me a lot of Apple and the way that I read this is, hey, if you don't get excited about my multiple beats and raises, then maybe this capital allocation and giving back to the shareholders is enough to spike. to spike the stock. And so far, it doesn't seem to be playing it. The other read item that I was looking at is VeraRax. Jensen said it's literally a multi-billion dollar business already. And I was very surprised for him to say that because we saw We saw Ian Buck on social media and he's chief hardware guy at NVIDIA personally dropping off some sleds to big customers, right? And I read that as first time they're dropping it off as opposed to a multi-billion dollar business. Final comment, company now breaks out their businesses separately down to AI, basically enterprise, sorry, data center and edge. And edge is exactly what you, you know, they're consolidating gaming, automotive, PC stuff and workstation into one place. I think that signals, right, they probably don't want to signal a deceleration in gaming, where they need to take wafers and packages from one to the other. And also the gaming market is getting hit big time because of the 2x price of memory. So all in all, market doesn't quite know what to do. company probably on paper, you could probably tell me better, it's probably worth $8 to $9 trillion, as opposed to $5 trillion.

DANIEL NEWMAN: 

You know, I watched your entire interview. I know you watch all of mine. And the smartest thing you said, and I really kind of recalibrated this as a victim of their own success is how I claimed it. But that people that have the chance right now to bet on credo and bet on a sterile labs that companies that can double and triple in a quarter in value. are looking at this $5 trillion market cap and saying, hey, no doubt this is a great company with great performance. But is it like the Apple of the last era? It's just long, safe money of continued growth and solid earnings and good profits. And we want to go all in on this exuberance and this melt-up. And we want to put our money into space and into quantum names that can go up 50% in a week on a single piece of news, even though no revenue has yet been created from it. It's the best company in the world right now, them or Google. I mean, arguably the best company in the world, unquestioned, $48 billion of free cash flow, amazing utilization of its cash for investments. It's still first leading with innovation, and behind its innovation comes investments into its ecosystem. People call that circular financing. It's idiotic because they're taking and getting in investments three, four generations before the market does. And they're getting their money's four or five X-ing as these companies continue to appreciate while utilizing their technology to build their future services. And so then on top of that, they 25X their dividend. They authorize the massive buyback, which maybe they use, maybe they don't. And their point is basically like, look, if the market doesn't appreciate what we're doing, we'll buy our stock cheap because we believe it'll be worth more later. This is a a company that will be 10, $15 trillion at some point in the future, unless they wildly screw up. And that is possible. I mean, look, you know that every generation, there's a company that has it by the, you know, they have it all in their hands and they screw it up, but I think they'll get it right. So those are great earnings. You know, I just, you know, what I kind of keep saying, sorry, my context all messed up. Just keep doing what you're doing, guys. Just keep doing what you're doing. Um, you know, next one, Pat, really interesting one. I mean, we don't cover this company super closely, but into it. Big beat. Okay. Um. But, oh wait, hold on. Earnings beat, revenue miss, if I'm right, and they announced a huge workforce cut. So this is going back to our SaaS argument. They didn't raise their guidance, and they authorized a buyback. So this is a really interesting one because this is a company that builds a lot, really well known for small business software, right? Small business SaaS. You run on it? No, you run on NetSuite, but I ran on Intuit for a long time. I think some of these cuts are really interesting. A lot of these companies are talking about expected growth, record results. We heard it from Cloudflare this week. We heard it from others. They're seeing a future of huge growth, but at the same time, they're massively cutting because they actually want to, one, you're seeing in some cases, incent their top performers. that are using AI and scaling more meaningfully. And in some cases, they're just basically coming flat out and saying it. They're just eliminating the context layer. People in the organization, they basically just reported and shared what happened. You know, a lot of these CEOs are coming out and saying, you're either building, you're selling, or you're delivering. And if you had any other role, these companies are seemingly just cutting out that fat. Any thoughts on that?

PATRICK MOORHEAD: 

Yeah, you know, I think first and foremost, I'm starting to get the feeling that these companies are cutting all these people to be able to afford tokens. And I don't know if you've thought about that, but it just, you know, as I piece together the expense warnings of tokens that we've seen from a bunch of companies and having these conversations and what these companies have to do, they can't keep, they have to cut expenses somewhere. This could be the case where they're just cutting people and figuring it out along the way. One thing we haven't seen is companies that have cut a ton of people and they've missed something. That's what I'm fascinated about. It's hard to cut 20% of your staff and immediately not have issues. We saw that with Intel. They cut five layers out of the organization and at least what it looks like for right now, haven't had a stumble. The other thing I find ironic is that this Intuit comes at a time that you have OpenAI, ChatGPT doing a finance essentially finance plugin. Where you can through, I think it's Stripe possibly, connect all your accounts up and you can query and you can do all this amazing analysis on it. Not that it's a direct competitor, but it's an intelligence layer that could sit on top of many of Intuit's products.

DANIEL NEWMAN: 

Yeah, absolutely. It'll be interesting thing to watch. Hey, Pat, let's let's bounce to another. You can call this I call these Nvidia tea leaves for those out there, because a lot of people don't necessarily follow Lenovo. But when you see Dell skyrocketing 15 percent a day, another company that's clearly on the rise and skyrocketing, part of why I think Dell's ripping too is Lenovo.

PATRICK MOORHEAD: 

Yeah, that's right. I mean, you saw beat on revenue, beat on, beat on EPS and, you know, a, a pretty good, I mean, gross margin beat by 1.6 points. And, and all, all I can think of, well, first of all, IDG was, and that's the PC, up 17%, likely a lot to do with price increases and Lenovo's ability to source memory and a lot of it out of China. ISG was up 30, 32%, not that 100% that we're seeing it. or 73% growth like we're seeing at a place like Dell, but Lenovo isn't necessarily hitting that gigantic Neo clouds. ISG, they returned a full year operating profit, which is really good. AI server pipeline, $21 billion, which is exceptional. SSG just keeps cruising along, increase of 19%. First ever year above $10 billion. I think that's a success story for the company. I do like that they're leaning into being very straightforward with AI mix. growing 84% year-on-year to 38% of the entire company revenue. So PC franchise, right? record share at this point. I don't know if that's from IDC or somebody else, but I'll take their word for it. So good stuff, Lenovo. You and I have a couple of pre-briefs or post-briefs coming up to get even more of the details.

DANIEL NEWMAN: 

Yeah. I mean, look, there's a lot you could look at, but where I'm You had the 84% AI-related revenue growth. That server, $21 billion server number, is incredibly strong. They're growing profit a lot. And of course, this is a company that has to make this pivot. They benefit just so much by having this part of the business because, of course, as PCs are looking to be down possibly double digits by the end of this year, this is where Lenovo, Dell have such an advantage to be able to have this part of the business to do a lot of augmentation and offsetting. It's great results and it's great for the ISG business and the solutions business will benefit from this mightily as well. One last one, quick one, Pat, we don't cover this one in depth, but of course you would presume that this would be a high performer and that's analog devices, record revenue, 37% year-on-year growth, top guide, 67% year-on-year growth of EPS, $3.09 a share, 73% gross margin, QQ3 guide crushes the, you know, 10% above, 3.9 billion versus 3.63. And of course, where's that coming from, Pat? Well, they're growing, actually, in fairness, across the board. Industrial up 56%, comms up 79%, and yes, data center up by 90%. They acquired Empower at $1.5 billion in the quarter. Another business that's just rocking and rolling. Some people say lagging or legacy or essential node chips are boring, but Pat, can you build servers without the analog devices and the lattices and all these companies that we talk about? Can you build physically all the things that we're going to build out? These guys play a part, right?

PATRICK MOORHEAD: 

Yeah. Nothing more to add other than you need lower power, you need analog. And that's where analog devices kicks in. Sure, they're in consumer, automotive, comms, industrial, but it was really the data center that in aerospace that, you know, kind of this national sovereign tailwind that's, you know, moving them forward. That was up 56% year on year.

DANIEL NEWMAN: 

And there we have it. Pat, we did it. Another one in the hopper. Episode 305, everybody. Appreciate you tuning in. I hope you like that new high-definition camera. See those bicep veins? That's really important. You know, that one sleeve of yours is showing a lot of gun, Pat. One sleeve up, one sleeve down. Everybody wants to see it. Everyone wants to know. They come for the tech, stay for the guns. Appreciate everybody tuning in this week, episode 305. Be part of our community, subscribe. See you all next week.

‍