On this episode of The Six Five Webcast, hosts Patrick Moorhead and Daniel Newman discuss the tech news stories that made headlines this week. The handpicked topics for this week are:

1. Meta Goes Community Notes, Eliminates Moderators
2. NVIDIA Outlines Its Next Big Strategic Opportunity Beyond Datacenter AI
3. NVIDIA Widens Enterprise Software Mote With Agents
4. NVIDIA Gets Into Developer Desktops: A Notebook Pipe-Cleaner?
5. Qualcomm Expands Auto Empire & Lowers PC Entry Price Points
6. AMD Goes Big With Dell & Targets Apple M4 With Max
7. Intel Reiterates Go-Forward Strategic Priorities

For a deeper dive into each topic, please click on the links above. Be sure to subscribe to The Six Five Webcast so you never miss an episode.

Watch the episode here:

Listen to the episode on your favorite streaming platform:

Disclaimer: The Six Five Webcast is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors and we ask that you do not treat us as such.

Transcript:

Patrick Moorhead: Welcome back. The Six Five is back after a small two and a half week sabbatical. We were on vacation. We may or may not have been together in the Snowmass, Aspen area. Dan, is your brain back to work?

Daniel Newman: Yeah. We didn’t really get a chance to shut down that much, did we? We would ski in, ski out and yes, we were together. I would actually play that off a little bit more if there weren’t a whole bunch of digital evidence of pictures of us eating dinner in a, I don’t know what that’s called, like a teepee?

Patrick Moorhead: A yurt.

Daniel Newman: In a yurt in the middle of Aspen, or all the chairlift photos, mountain photos. It was a-

Patrick Moorhead: Steak photos. Half of our photos were steaks, I think.

Daniel Newman: It was a bestie fest.

Patrick Moorhead: Totally.

Daniel Newman: It was a bestie fest. And we hung out with Bezos a bit. Not really, but we were close enough to almost see him that one time.

Patrick Moorhead: Yeah. There was a bunch of superstars. J.Lo, we ran across in the Ralph Lauren store. There was a JD Vance spotting-

Daniel Newman: On the mountain.

Patrick Moorhead: … out there as well. And all the Kushners were in Snowmass.

Daniel Newman: Wasn’t Kurt Russell? For the older folk over at Kemo Sabe getting a new cowboy hat.

Patrick Moorhead: Yeah. Yeah. And I got a ridiculous picture of him. It looks like he’s smiling at me. I’m pretty sure he was smiling at Goldie Hawn, who was carrying a big huge hat bag. But we’re back and we started on Sunday night, you and I both flew to Las Vegas to attend the Consumer Electronics Show 2025. And I think we have a few topics related to that. So, what are we going to talk about? We’re going to talk about Meta going all community notes, eliminating most of their 14,000 moderators. We’re going to talk about the next big NVIDIA move beyond data center AI. They also announced some new enterprise software, which may or may not be widening their enterprise moat. NVIDIA also got into a developer desktop, but is it the notebook pipe cleaner? We’re going to see.

Qualcomm won a bunch of awards at the show, too many to talk about. But we’re going to focus on what they did to expand their auto empire, and also what they’re doing with the price points on their Copilot+ PC processors. And then we’re going to round it out with some AMD discussion, a big deal with Dell. They’re now targeting Apple’s M4 processors. And then we’re going to wind up with another chip company, Intel. We’re going to talk about what we found out related to their strategic priorities with the new leadership team, the Dynamic Duo. So, we’re going to kick this off with Meta, big announcement. The press is going absolutely bonkers. Essentially Zuckerberg and his three social media properties is going to go all community notes, very similar to X. Dan, is this the end of the world for everybody? Are we going to be crying memes out there?

Daniel Newman: There will be some crying memes, but Pat, let’s face it, there are always crying memes. When there’s too much moderation, people think that they’re having their free speech taken away. When they have too much free speech, people think it’s all hate speech. When I can’t tell you, Pat, that you still got food on your chin right now, it hurts your feelings. There’s a feeling hurting barometer that goes off and the expectation is you need a moderator to take that down. And then on top of it, there’s just the business and economics and politics. What we saw in November was a, we can span whether it was decisive, whether it was a landslide, whether it was a close defeat. There’s a lot of different metrics that could be used and describe what happened in the election. But what happened was there was a feeling that what would be considered to be strongly left or liberal politics needed to come back towards the center in some ways. I know, we’re not a politics show, so just put up with my crap and listen to me for a minute.

This is why you have Tim Apple, Tim Cook, Mark Zuckerberg, and basically every other CEO flying to Mar-a-Lago right now. Or if they’re invited and they’re writing their million dollar checks for the inauguration. Because the bottom line is, most of us feel that the politics that these folks articulate is some permanent state of their beliefs. But realistically, it’s a state of the climate of the economy and the world and the business, and their goal is to continue growing their companies in every different regime. I know that’s not Meta exactly, but let’s face what really brought the whole community notes and moderation to the forefront. Really goes back to some of the things that happened with early utilization of social platforms to drive election or opinion at the beginning of misinformation. This didn’t start in the Trump era. This goes back to the elections with Obama.

Patrick Moorhead: 2016.

Daniel Newman: What’s that?

Patrick Moorhead: It was 2016, I think.

Daniel Newman: That’s when it really started, its first Trump election. But it had been happening for some time. We got to this era where people figured out that they could use social media to move opinion. And by the way, people are not particularly able to discern between fact and fiction. They’re not able to always completely understand the nuance of information. We’ve seen how different outlets, media, different publications, this can span both of the sides of the political continuum, can take snippets of comments that are made out of context. I still think the best example of this is that Trump calling the far right extremists, very fine people. And that was debunked everywhere, including on Snopes. But I had an argument with my mother about this in the most recent election, because she still to this day believes he said that.

So, this is what’s going on. And what’s happening now is there’s still this bifurcated view of the world that X is a niche play, but it’s really become the number one news source for, I think Musk shared the other day, over 150 countries. It’s the number one news source. It is the first place, it is the fire hose of where people get information from all major media outlets. But also from these highly reliable citizen journalists that are out there that people trust, whether that’s you Patrick, about fitness and healthcare, or other people about technology. But in all seriousness, Zuck went the right way. And look, I do think there is a place for moderation, meaning when there is violent, horrific content, child exploitation that takes place. I think that is a real meaningful thing where maybe moderation should take it down before it becomes harmful and gets into the wrong hands.

Having said that, an opinion about healthcare, an opinion about a political bill of law, an opinion about an economic decision, an opinion about a political figure. We’ve seen over and over, Pat, on both sides of the political divide, information has come out, the media has basically done some type of censoring to provide the belief to the public that a story is wrong. And then we found out that that story turned out to be true. It’s happened a lot. And so the better way to do this is using real time community notes from sources that have to be validated, that have to be checked and have to be willing to be public about what the sources are, who the sources are. And that way people can read it, the community note appears. I don’t know about you, Pat, but every time I see a community note, I instantly go, well, maybe I shouldn’t have shared that, or maybe that is inaccurate.

So it works for me. I think it can work for others. So, I think Zuck is following the momentum. He understands that this country is politically divided, but that there are two very large swaths of potential consumers for the platform. There will always be people at the farthest edges of the fringes that are going to feel that they are not doing enough to support their political beliefs. But just like the way primaries have destroyed political parties, too much moderation can destroy social platforms. And we need more centrism. We need more town square, we need more debate in public, Pat, and I think this is a good thing. And I also think it’s good that he’s moving what’s left of the moderation to the great state of Texas. I believe I also heard that that’s happening.

Patrick Moorhead: Yeah. Good rundown there, Daniel. And really is a shift for Meta and whether it’s politically motivated or something different, it seems like we are moving to more of a centrist area. And just to be clear, it’s not removing all humans from the loop. In fact, the way to get reinstated, they’re actually increasing the amount of resources if you get banned or things like shadow-banned. And I think the biggest thing that came out, and I do appreciate the opacity of the blog. Zuck came out and said, “Hey, we’re going to loosen restrictions on some sensitive topics here.” As it relates to community notes, it’s just not one way or another. It actually, how many times has Elon been community noted?

Probably 5%, 10% of his posts get community noted, and I think that just shows that the system is working as it should. We will, and we’re already seeing examples in press stories about the type of content and how horrible, but I think I’d like for everybody to give it a chance. I also really like the focus on the high severity violations, like terrorism, things like child exploitation, stuff that I’m hoping that we can’t have a disagreement on. But all in all, I’m an old guy as you like to point out, and the pendulum always swings, and the pendulum is currently swinging in a different direction. So, I like X because of the freedom. That’s a personal statement. I don’t do a lot on Facebook, or Instagram, or I forget the text-based platform that I get bugged on, on Instagram.

Daniel Newman: Threads.

Patrick Moorhead: Yeah. But maybe this will give me a reason to go back to Threads.

Daniel Newman: And Pat, by the way, I don’t think, we have so much to talk about that we can’t spend too much more time on this one. But we are, I think today, we’re very close to a TikTok ban, so Zuck obviously sees a fairly significant reason to potentially try to win over more audience. That’s a big opportunity out there. I do want to say though, that watch he wore really showed that he’s a man of the people and I appreciate that.

Patrick Moorhead: Way to round it out and we can’t spend more time.

Daniel Newman: $900,000 watch dude, that’s epic. And by the way, such a baller move. We’re in the era where you could do that again.

Patrick Moorhead: Yeah. And the next thing he might do is show up on Rogan and go through this. That’d be crazy. So, let’s move to the next topic. So, NVIDIA stock is rocking because of what it’s doing in data center AI. But believe it or not, it’s actually still very engaged in elements like automotive, the industrial edge, and of course gaming. Dan, you and I attended CES where we had a couple engagements there at NVIDIA. I was fortunate enough to have a conversation with Jensen. It wasn’t scheduled, but it was very happenstance. I’ve known the guy since 1998, but you and I were both part of the Financial Analyst Day and the Q&A. And I think we both watched the keynote in the comfort of our own rooms. So, what is the next opportunity? And the company has been laying out cookie crumbs for a long time, and particularly on its financial analyst calls, in the amount of time it devotes to topics that are not data center AI.

So, this notion of primarily robots, we know that they’re in the automotive industry, but industrial robotics is the next big thing. And when I say industrial, that’s a loose term that can go even into healthcare. But essentially they introduced a product called Cosmos. And think of Cosmos as a world foundational model to be able to essentially train robots. And they’ve got different levels. And you might be asking, “Well, wait a second, isn’t that what Omniverse did?” Well, think of Omniverse as having very, very detailed and precise models in the physical world. It’s a little bit more of a challenge though to extend that into the virtual world and to be able to train to essentially make these robots a lot smarter. So, if you take Omniverse and you feed that data into Cosmos, the theory here is that the quality goes up.

And we’ve seen synthetic video worlds being created before. NVIDIA did it with Omniverse. And you see companies, like OpenAI with Sora, and you see companies like Adobe, they haven’t gone GA on this, but video Firefly to be able to create synthetic. And the challenge though has been if you’re trying to train a robot with a high degree of accuracy in what it does literally and how it interacts, the quality of the video output has to be higher. So, it’s clear that the next big planned growth area for NVIDIA is in robotics. You can use Cosmos for automotive. We’re going to see this, and let my imagination run wild, why wouldn’t we use Cosmos for something like entertainment, right? And the ability to be able to improve the output of entertainment content, basically a better Sora.

Daniel Newman: Is that it? Are you done?

Patrick Moorhead: Are you there, Dan? Tap the microphone. You there?

Daniel Newman: No, I was there, I was just-

Patrick Moorhead: Stock portfolio.

Daniel Newman: I’m not used to you talking that little. I’m just like, Lord, that was actually really fast. Pat probably has another workout to get to. When you’re getting huge, you just don’t have that much time for this stuff. No, in all seriousness, that was really a good breakdown. I’m going to try to tie some threads together. I know we have a seven five today with three NVIDIA tracks, and I know they’re different topics, but some of this is all the same to me. And so the thesis about NVIDIA has been that its business is basically valued for the number of data center GPUs it can sell, and then of course the attached software licensing. And it’s all really been attached to the amount of expected CapEx spent by the largest hyperscalers in the world. The antithesis of the business has been whether or not the hyperscalers of the world are going to continue to build all of the AI in the future on NVIDIA’s systems. Or are they going to vertically integrate with their own GPUs built with Broadcom or Bell? And by the way, these things are somewhat interdependent, but in some ways mutually exclusive, because I think the answer is going to be both.

Having said that, I said the entire CES keynote was all about NVIDIA showing that their AI extends beyond the data center, and here’s all the ways it extends beyond the data center. But, and it’s an important but, at the same time, all of this will create more demand in the data center. Because let’s face it, $30 $40,000 GPUs, sold 72 at a time with hundreds of CPU cores and licenses attached to that. That ultimately end up then creating more licenses for Omniverse, and then more licenses for edge based solutions and Orin in the vehicles and everything else that’s going on is a really, really lucrative business. And so the themes of this event, and I’m going to stay with the one theme here, but there were three. The three themes were, it was physical AI, it was automotive, and it was PCs and future of client. So, you hear basically talking about this physical AI thing. And the physical AI thing as I see it, is the next trillion dollar plus leg up for NVIDIA. This is where they get the next trillion dollars in market cap. Why? Because two things.
One, it’s really hard.

So, all the things that you’re hearing about ASICs and accelerators for the data center, what NVIDIA is doing here, none of that stuff can do that right now. Now they will be able to build that stuff. But the second thing is how this all threads together between that simulated world with synthetic data and the physical world with real data. We heard Elon Musk sees Tesla stock trade at something like five times higher than the average mag seven in terms of forward multiple. And the reason is, is because people believe he’s going to solve two major themes of the future around AI. One, autonomy and driving, which by the way, Waymo does better right now, but he has more cars, more data. There’s a lot of reasons why Tesla could win. The second thing is he’s going to solve humanoid robots. So he’s got, you remember the guy on stage in the suit two years ago, and then this year they had the ones pouring drinks that were semi, I think there was someone in the back probably with a remote control doing things. But we went to CES this year, you saw physical robots.

You saw robots that could move boxes and robots that could walk around with us, and you saw robots that could offer companionship. But the thing is, is that if you just get to the level of every human has a robot, just that, and you’re talking about something that costs as much as a Tesla 2 or 3 at 30 to 40,000, the economics of this become really powerful. Well, someone actually has to build the software stack to train that robot to be useful. Otherwise, you end up with the little things we have now that walk around the house and then fall over and set themselves on fire. The thing that walks around your pool that forgets to clean. There’s robots now. There’s robots in commerce, it’s everywhere. But Jensen is basically saying the mechanics of the human, the physical mechanics of the way we move, say walking up a hill or down a hill, is really hard. To simulate that you would need to literally have tons and tons of training data.

Well, with what they’ve built in the connective tissue between Cosmos, GR00T, Omniverse, you can basically simulate this stuff faster, build models, and then get us to humanoid robots a heck of a lot faster. So, it’s really cool, exciting stuff, Pat. And so I know we’re going to talk more NVIDIA, so I’ll end that theme on that one. But I really think the market misunderstood what happened. I think it was basically, and he said the ChatGPT moment for physical AI, but I just think it’s the stack. The stack that’s become the all in data center, digital AI stack that people are using for software. We’re seeing them build something very similar for the physical world, and that’s going to be really hard to replicate and it could be a really great long term moat for NVIDIA.

Patrick Moorhead: Good stuff, Dan. And by the way, one other thing, big announcement that came out of CES was NVIDIA announced yet another enterprise software stack that sits right on top of NeMo related to agents. Dan, what is it, and what’s the significance of it?

Daniel Newman: So, basically agentic AI for all of you out there, there’s these two schools of agentic. And so I’m going to talk about it a little bit theoretically here and, Pat, maybe you can add some color as well on what NVIDIA is specifically doing as well. But this has been the hottest trend in digital. So, we saw this evolution of, two years ago, the moment was about generative, text-based, and then you saw multi-model, multimodal, text, speech, video. And then you started to see smaller language models evolving, which is text language model for very specific use cases. And then you’re seeing the companies like ServiceNow, Salesforce, Microsoft building the next wave, which was assistance, which are basically AIs that could do a specific task.

And now you’re seeing agents, which is the idea of being able to basically deploy a force of assistance that can work on various tasks and actually co-communicate. They can intercommunicate between one another. And so you could essentially put forward someone to do a work stream on supply chain, or a work stream on operations, or a work stream on finance. And you can say, “We’re working on these 3, 4, 5 work streams,” and all the agents can work together. So, what needs to happen for this to work is there’s a lot of orchestration that is required. And so there’s these two consumption layers that are going to be really important for people out there. There’s the software ISV consumption, which is what we’ve heard a lot about.

So just use Agentforce as an example. Agentforce, the idea is you’re a Salesforce customer, you can get pre-built agentic AI that can then be put into your environment that can do sales motions. That’s very cool. Well, guess what? Someone has to develop all that stuff. It actually has to be developed and Salesforce’s platform itself doesn’t really develop agents. What happens is they train agents on a platform like what NVIDIA is building with Blueprint, that can basically take a bunch of pre-canned agentic capabilities, mix it with proprietary data sets, put it all together so then it can be deployed in software. And NVIDIA has been left out of the overall agentic conversation, believe it or not, despite the fact that they’re the hardware for training everything. You really didn’t hear about agents and NVIDIA until this event. And so I think that’s in my take, that agent and Blueprints to NVIDIA is for the developer what the ServiceNow Microsoft and Salesforce products are for the user, for the enterprise. Someone has to build the agents, NVIDIA is trying to make that easier.

Patrick Moorhead: Yeah. It’s pretty wild the amount of software that NVIDIA is generating. And I’ve been super disappointed with the enterprise, I’ll call it the on-prem enterprise capabilities to be able to do that. Really nobody has stepped forward yet. I think the closest example of it are bits and pieces of things that, let’s say IBM has done with their Watsonx platform. You’ve seen a little bit of the data management action from companies like Cloudera, but pretty much a little bit of model management with Cohere, but there hasn’t been a lot of action here. So, one thing that Jensen had said or reinforced, I’d heard this before, he reinforced, “Hey, we’re market makers, not market takers.” And I think this is a market making opportunity. And they had put forward different agents; a research assistant, a financial analyst, employee support, customer service, sales development, all the ones that Daniel, you and I have seen in our travels and travails to these enterprise, SaaS companies. And if they can light these things up for the enterprise, I think they’re shrinking the time to market that looks like a couple of years to really scale enterprise AI. And because they’re NIMs, you can run them anywhere. You want to run them in the public cloud, do it. You want to run it in your private cloud, run it. So, AI Blueprints sit on top of NVIDIA NeMo that I know it’s not per se a data management platform, but you’re evaluating, you’re customizing, you’re curating, there’s guardrails and it also includes a RAG capability on top of that.

Daniel Newman: And by the way, they’re doing a lot of interesting stuff with models, the Nemotrons, what they’re doing with Meta, Pat. You could say it’s not a data management platform, but I would argue how long do we need data management platforms in an era where AI should be able to discern different data format types? I don’t want to get too far ahead of ourselves. By the way, one other thing, I just got an email literally as we’re talking from NVIDIA about their AI Blueprint for retail shopping assistance. That’s the example though, is when you hear the stories from a Salesforce about what they’re doing with Nordstrom, someone has to actually build and train a model that can then be implemented into the SaaS platform.

Patrick Moorhead: That’s right.

Daniel Newman: That’s the TLDR here, right?

Patrick Moorhead: It is, it is. And it’s funny, I don’t applaud NVIDIA on a ton, but I think this is great because quite frankly, there’s nobody else doing this that you can apply this to on-prem capability. And Daniel, I have not talked to a single CIO or a single CDO that says, “Now, I’m going wholesale into the public cloud.” They want their strategic data for the most part to sit there. They’re afraid of the cost of the public cloud. Not everybody of course. Just look at the growth that the public cloud is experiencing, but I think this is a very positive move. So, let’s move from software and agentic services to NVIDIA getting into desktop. So, NVIDIA is not new to creating devices. In fact, nearly 15, 20 years ago, even when I was at AMD, they did a foldable chip for WinRT.

It was ARM-based, had a GPU on it, along with Qualcomm and Texas Instruments. And then they’ve also done these gigantic desk-side developer platforms as well. I believe it was a desktop desk-side DGX. It was a monster though. And what they introduced at the show was something called Project DIGITS. Think of it as a very compact desktop. And not too many things surprise me, but this one did. I knew because people are telling me, not NVIDIA, but everybody else, that they’re going to be bringing out a notebook processor that will have devices in it, the first half of ’26. So, when it relates to it being a, I like to call it a pipe cleaner, I really find this fascinating. The size of it is crazy. It’s literally like this big, which I-

Daniel Newman: It’s like a Mac mini.

Patrick Moorhead: Yeah. And it just has the coolest-looking design too. It looks like it fits in a 1U rack. Obviously it doesn’t, but that’s the feeling that you get. Has a new chip called the GB10. Not to be confused with the GB100, and GB is Grace Blackwell, but they partnered with MediaTek to make this happen. And I think they had to partner with MediaTek because low power is really not what NVIDIA is great at, and the ability to churn and burn an SoC in very quick time. We’ve seen this with Qualcomm. Their ability to churn out something so quickly is absolutely amazing. And the performance is really, we really don’t have a sense even though we know it has 20 ARM cores, and this is the new Blackhawk core that I wrote about last year. But we do know its AI performance, which are just absolutely kind of mind-blowing. I think it’s half of the performance of ADA out there at 500 teraflops, which is seriously, and that’s FP4 performance, which is absolutely insane.

So, priced at $3,000, available May of 2025. In the limited time that I had to talk with Jensen, I think the first thing I asked him is, “Hey, can I get Windows on this thing?” And I’ll note that there was a lot of confusion based on what Jensen said. So, this ships with Linux and not Windows. And I think when he talked about this on stage, he did not get his message across very well. And you saw Satya Nadella weigh in as we saw Pavan on X trying to clarify what was going on here. But essentially Windows has a Linux subsystem, and the TLDR is all the great software that you can run on DIGITS, because it comes with the full NVIDIA stack, you can run on Windows in what’s called WSL, and that is it. NVIDIA is not abandoning Windows and moving everything to Linux, but their software stack, their entire software stack is Linux optimized, because it’s for the data center.

Daniel Newman: And there you have it. I think some of it’s pretty provocative, Pat, and I am with you. I don’t see any massive immediate changes. Like I say, the end of enterprise SaaS is going to lead to these deprecated services and databases and abstractions and new toys. There might be a future in which a Linux Kernel and a chat interface is all you really get when you open the lid. Because you certainly aren’t going to see apps on your phone the same way in the long term, if agents actually do what they’re supposed to do. That’s the bottom line. If they actually do what they’re supposed to do. So it’s funny because all these companies are in conflict with themselves. It’s like they’re creating their own demise, but at the same time, if they don’t, they’re going to miss the trend and become the Kodaks and Blackberries of the next era of pundit stupidity. Anyway, I think you hit that one pretty good. I don’t have a lot to add to it. I think I did like the launch. I thought it was cool how he came out and took the big GB box and made it small. I think the way he told the story, I think the idea of having a high-powered AI machine at your desktop for developers, super, super-duper cool. I do want to just say before we leave the topic of NVIDIA and CES, that I did find it particularly funny that the CEO of CES kept calling it NVIDIA.

Patrick Moorhead: I was wondering if you’re going to bring that up. Gary Shapiro.

Daniel Newman: Yeah. The most consequential company of our time, NVIDIA. All right. I’m not going to say whether that really bothers me or not. I’m just going to say that if it is the most consequential company of its time, we should learn how to say its name. So, for everyone out there that’s listening that’s been unsure of how to say it, Pat, how do you say it?

Patrick Moorhead: NVIDIA.

Daniel Newman: Thank you. I didn’t want to say it wrong, so I put it back in your lap.

Patrick Moorhead: I appreciate that.

Daniel Newman: All right. That’s it. Let’s keep going. I don’t need, you nailed DIGITs.

Patrick Moorhead: Let’s get into Qualcomm. So, Qualcomm came in, did seven or eight announcements. They were primarily around automotive and an extension of their PC, SoCs, Dan.

Daniel Newman: Well, first of all, it’s really interesting, Pat. A year ago, you were on stage with a whole bunch of the executives of the OEMs on the devices side. It was the year of the AI PC. Being self-critical and going back in a year of review, I never said emphatically that we were going to have a super cycle, but I think I was probably more optimistic about how the year would go than it went. I will mostly blame software for being wrong because if software roadmaps had met what I’d expected, I think there would’ve been more allure to people buying AI PCs. The reason I’m pointing this out while we talk about Qualcomm is Qualcomm was obviously the new, exciting, disruptive, ARM-based entrant that was going to create the new future of these Copilot+ PCs. And the year in review is you ended up with a really exciting silicon year, a race and an ARMs between Intel, ARM-based Qualcomm, AMD, and now to a lesser extent, MediaTek and NVIDIA, all ripping and racing towards this era.

But here’s the question and the point, Pat, is Qualcomm comes out with a slightly lower cost, a mid-tier, to address probably a subpart of the market that their early Elite product was going to be too expensive for. Part of which is because we’re getting into this era of AI PCs or Copilot+ PCs. Part of it is because what’s ended up happening, at least so far, because of maybe software not quite meeting the moment yet, yet is an important word here, is we’ve ended up with needing to address price points again to be able to attract a bigger market to try to accelerate updates. Because what we do have, Pat, are really powerful next generation devices with really good battery life. We did end up getting that. And in the process we ended up getting also some of the best out of, and we’ll talk more about Intel later, but some of the best out of some of the chip makers to really get back to competing. Because in the beginning, Pat, it looked like Qualcomm was just going to run away in terms of specs.

And now we’ve got a three-headed race and we’ve got some really competitive parts, but the new X product is going to hit that middle price point, Pat. And I think that’s all about accelerating people moving over and buying in and utilizing, because I still think there’s a lot of questions on what software is going to drive people and how much on-device stuff happens in terms of it being a value add for the consumer.I get why cloud companies want it. It’s expensive to have all these things being done in the cloud, but do the buyers notice the latency and care about having to go to the cloud to do a query on ChatGPT, Perplexity, or any other language model? I think that’s something we got to discuss, Pat, and we’ll spend some more time talking about that, I’m sure over the course of the year.On the other side of this, Qualcomm has been an absolutely amazing journey in automotive, and the announcements keep ticking. And while we did hear NVIDIA seem to have a more bullish forecast, Intel has talked a little bit about some automotive progress that it’s making. Qualcomm has been over the last, I’d say three or four years, the de facto or undisputed winner of the automotive opportunity building a nearly $50 billion automotive pipeline. Business is growing every quarter. They’ve had more design wins.

And here at CES, they just furthered partnerships with Google, partnerships with Garmin for GPS and Unified Cabin. With Tata, which of course is a big India opportunity for the company and in other markets. Partnerships with Alpine, Amazon, Leap, Mahindra, I’m just reading this by the way, Hyundai, Royal Enfield all doing more with AI-powered Cabin. They have a new partnership with Panasonic for the Cockpit.And Pat, they’re also showing more and more how generative AI can work in the vehicle. And of course we know you can’t have full language models, and there’s a lot of work to be done to synthesize these models down to make them power efficient enough to be run at the edge. So, this is something Qualcomm is well-equipped for. The customers, the wins, the designs, the partnerships indicate its success. It’s continued progress as I see it. It was very a high volume of what I call small movers, that all show Qualcomm is moving in the right direction.

Patrick Moorhead: So, I like to look through old photos that I took at shows years ago, and literally four to five years ago, I didn’t know that Qualcomm was even looking at ADAS. And they asked me if I wanted to take a ride in their ADAS-infused car, and that was a huge surprise to me. And here we are, the backlog stands at $45 billion and in their Financial Analyst Day, they put 4 billion in automotive revenue by fiscal 2026. And then doubling that to 8 billion fiscal ’29, really driven by ADAS and even fully autonomous driving. And their announcements here were really just, I’ll call them tuck-in announcements, this Snapdragon Ride Flex SoC with Hyundai, important movement. It’ll be interesting to see on the software side, Qualcomm versus NVIDIA and what they can do. NVIDIA has made most of their moves, I would say in the autonomous driving. Qualcomm has taken most of NVIDIA’s dashboard design wins, and Qualcomm has always been a huge player. And obviously the connectivity, whether it’s 3G, LTE, or 5G inside of the car.

So, strategically, the company just gets moving forward here. One thing I want to focus here is, is on this new Snapdragon X. With adding X, now there’s three tiers. You go all the way from $600 to whatever, $2,000 with the entire Qualcomm Snapdragon platform. It’s going to be really important for Qualcomm this year to do a few things. So, first and foremost, they really have to nail education as it relates to performance. And what I mean by that is, is Qualcomm doesn’t win all the performance figures all the time, but what they do do is they win most of the speed races when it’s not plugged in, like you’d expect a notebook to do. Can they punch that through there? Qualcomm has to scale. It was a challenge to do that before because of a couple of things going on. First of all, price points. But now as low as 599, you’re going to see platforms out there. You also have a combination of desktops and notebooks that will be coming out. I did say last year that I expected the Super Sky goal to start sometime in the second half. What I didn’t factor in is that Microsoft would miss their software schedules of their Premier AI feature by eight months. So, Recall was supposed to be out May, June, GA, it is still not.

It is a very useful tool out there that I use every day and I think it’s pretty good, but it was severely delayed. And the second thing is, is to start getting traction inside of the enterprises to take advantage of the Windows 10 to Windows 11 transition. Qualcomm is investing a tremendous amount of money in there, but it’s going to be tough. AMD and Intel are running serious plays in the enterprise and don’t want to give up any real estate whatsoever. You have Intel playing defense and AMD playing hardcore offense. So, anyways, good stuff from Qualcomm. So, let’s talk a little bit about AMD and some of the announcements here. They made a ton of announcements. What? Is this your topic? Huh?

Daniel Newman: No, it’s all you buddy. Don’t let these gorgeous pearly whites distract you. Just keep going. I’m just smiling at my bestie.

Patrick Moorhead: Sorry, I just thought you were looking at me and laughing. That’s okay. So, AMD made a lot of announcements, but I think the two that we thought were the most impactful was they signed a go big. They announced a go big deal with Dell on their commercial lines that cut across desktops and notebooks. And it’s interesting, we’ve seen a lot over the years, I’ve seen many announcements. I’ve been involved in deals with Dell and really up to this point, Dell has not been able to effectively drive volume of AMD equipment, AMD desktops and notebooks. And it’s not just a Dell issue. AMD did not invest enough in the channel. I still don’t think they’re investing enough in content and programs. AMD really is about best product wins, but when it comes to enterprise desktops, that’s not just the case. That’s not the case. You’re not going to get somebody to pivot and go with you because your platform is $50 or $100 dollars less because it’s really a TCO. And quite frankly, people don’t want to change if they’ve been satisfied with what they’re using.

So, moving forward, I am hopeful that, and I’ve seen, I’ve talked to both companies, all the executives here, I’ve looked at how the two companies are positioning it. I’m looking forward to get a even deeper dive in the go-to market with both AMD and Dell. And made AMD stock rock, which good for them, but they’ve got to turn this into a reality here. Final thing I want to talk about is AMD brought out with what I think is a super differentiated product called Ryzen AI Max, lovingly referred to as Strix Halo. It’s the top of the line AI processor with 50 TOPS, but it’s not necessarily the TOPS. It has a big GPU maybe up there with the RTX 4060, a unique memory, a shared memory system. And it comes at a reasonable TDP between, I would say 45 watts is reasonable, but it goes all the way up to 120. This is a multitasking monster.And as you would expect, AMD outlined a lot of use cases related to content creation, which Apple has tended to dominate as of late. And I like that AMD went right after the M4 on this. So, probably the most differentiated product that AMD brought out the entire time. I believe that they’re partnering, at least publicly now, with HP and ASUS. I want one of these by the way, in a desktop format.

Daniel Newman: I want you to have one. I’m going to buy you one.

Patrick Moorhead: Thank you.

Daniel Newman: Merry holidays.

Patrick Moorhead: Thanks dude.

Daniel Newman: All right. Look, I will just add to this that I think it was a really big signal of intent that both AMD and Dell are finding a way to expand and grow that partnership. I think for AMD it’s an important growth area, especially in client. I like the work that Dell is doing to simplify its nomenclature or its design. Pat, you and I are both, I’m becoming a product marketing guy as I spend more time around you. And what I would say is I felt for a long time that Dell’s notebook lineups were really hard to discern and to understand, unless you were paying attention all the time. I think that Dell-AMD, this was probably the biggest go forward, I think the two have had so far in this space. And so I think it was a really strong moment for the company. In terms of what’s going on, I would also add that I think Dell, they did a bigger splash with Qualcomm when the first wave, they did that at Dell Tech World. We’re going to talk about Intel in a second, but let’s be really candid about it. Look, there’s some hope for Intel on the product side.

I’ve got some positive things I’ll say about it, but I also think if you’re Dell or you’re any of these companies, this is the hedge. You have to be hedging right now. There’s so much uncertainty. I think a lot of good work is being done to shore up some of the risk by Intel. But gosh, in AMD, Dell needs a stronger partnership. Dell and Qualcomm needs a partnership because until we really know for sure that Intel is out of the woods, these companies would be crazy to not be diversifying a little bit. And I think this was a signal of that intent. But at the same time, everyone out there knows Dell and Intel are like, and there’s a huge and very tight, very long and very deep partnership. So, we’re going to have to sit back a little bit, Pat, and wait and see if the numbers do the talking.

Patrick Moorhead: Let’s dive into our final topic. You and I both met with MJ at Intel, one of the co-CEOs. MJ is running products. We also attended Intel’s press and analyst conference as well. What were your takeaways, Dan?

Daniel Newman: Yeah. The conference was, it was a iterative moment of what’s going on. If you wanted to do the check, check, check, it was like Lunar doing really well, Arrow not as bad as people were saying. I think that was the takeaway is that there’s some of the kinks in Arrow Lake are being worked out. There’s a lot of focus on Panther Lake and what’s going to happen there. That seems to be a big moment for Intel, and that’s really the big ATNA moment that people are going to wait and see at a high volume part. But also, Pat, I think you said this earlier, maybe you said it online, maybe you said it offline. But I also think it was a really important moment for the company to refocus that its product first, process second. Which to me by the way, and I’m not starting any rumors, I don’t have any inside information. I’ve talked to the executives there for a long time. I think the separation/ or spin is getting closer and closer. I just think what I felt from that event was product, product, product. This company is going to get back to its roots of strength and product.

I also think it was pretty willing to acknowledge that a lot of its pain has less to do with the client and more to do with data center. The fact that it just doesn’t have a competitive AI part right now is just absolutely killing the company. Has to get that going. It has to get DCAI back on track. So, what I said in that last segment, I’m just going to weigh in a little bit more on it and then I’m going to pass this to you, because I’m also getting close to time and I want you to have some moments here. I got corporate BS meetings, everybody. Don’t you know how many corporate BS meetings I got to go to? So, I got to end this podcast with you. So much corporate crap, I can’t-

Patrick Moorhead: So sad.

Daniel Newman: When I’m old Pat, like you, I’m just going to pod all day. That’s all I want to do. Just podcast and crap talk on Twitter. But the other thing I would say is that the client’s side of the business has shored up. It just isn’t that bad. I think that they’ve made the big investments. They’ve kept the biggest customers in tow in many ways. They haven’t seeded as much market share as people have feared. The ARM stuff has done okay, but it hasn’t crushed Intel in any way more than Intel was already being crushed by its own challenges and problems. And I think MJ got up there and reiterated that a little bit, but I still say, Pat, until they get something figured out competitive in the data center beyond Xeon, it’s always going to be a problem for the product side. And I already said my piece about the foundry side. I really have a suspicion and a hunch there, but I’m going to let that play out.

Patrick Moorhead: So, from a priority standpoint, strategic priorities, this is really a big carry through all the way from the themes of Pat’s departure, the discussion that both you and I had with the co-CEOs. It’s subtle, but it’s important. Overall, for Intel to fully succeed, it has to have competitive products, competitive process, competitive foundry. And my thought was that 75% of the focus before this was on foundry, and now its products. Which it’s interesting, I think it’d be very hard for the company to have competitive products without a competitive process. But if you’re using somebody like TSMC, you could have much more competitive products.And I think this comes down to resource allocation and focus of the senior leadership team. And it seems like a subtle change here, but this is, I think Pat probably exited the company based on its competitive posture with its products, particularly on the data center side. Xeon has a lot of highlights, but AMD still continues to gain market share with the hyperscaler data centers, and it’s nearing double-digit with the enterprises as well.

And interestingly enough, doing really well with Dell on the EPIC side. But I think that the one thing that nobody can dispel is that Intel missed the boat on a data center AI GPU. But somehow in the same period, AMD was able to turn their 64 bit high precision FLOPs monster into a very competitive inference with low precision TOPS. And Intel will be coming out one to two years after AMD on this, and that’s probably a fireable offense, I guess, or I’m sure added to the momentum for a management change. Listen, the proof is in the pudding. I do believe that MJ is going to be listening more to customers, and she already has that reputation on the client side. Intel is a very large organization, where we see how quickly they can get this out. One thing that I find super interesting is, is that I do believe that the marketing and sales tactics are going to shift more back to where I remember Intel being strong. I think that Intel is going to price their products to not lose any market share, particularly on the client side. We will see. So, Dan, great show. Thanks for going over. I know that your corporate meetings are important to you now.

Daniel Newman: Well, you going to go to the gym again? What are you going to do the protein shake?

Patrick Moorhead: It’s super important. I’ve got six advisory calls.

Daniel Newman: You going to eat some eggs this morning?

Patrick Moorhead: Oh no, I’m absolutely going to go at steak eggs and avocado.

Daniel Newman: I just want to say this, I’m going to actually stay one more minute to say this to everybody. I walked around CES with you for three days. I had 50 strangers come off the street just to tap you on the shoulder and tell you how inspiring you have been. While I’ve had the joy of spending the entire process with you and being inspired by you, it was very inspiring and aspirational for me to see the difference that you can make in people’s lives. So, hopefully for everyone out there, we make a difference in the lives of the way they see tech, the way they think about what they buy, what they purchase. We certainly don’t give stock advice, so don’t ever make any decisions based on that. But Pat, way to keep crushing it, buddy. I’m going to drop.

Patrick Moorhead: Great show, Dan. Thanks for the thanks for the kind words. But listen, you were the guy who I was watching who could figure out how to stay healthy on the road. And so I think there’s a back at you bestie. When you used to tell me you would get up at 5:00 to go work out at 6:00, I thought you were nuts. And here we are both in the gym at 6:00 AM. By the way, we’d be there at 5:30 if it opened earlier in Las Vegas. But hey, thanks everybody for tuning in. We’re looking forward to a great year. We’re going to try to shift a little bit of the show. We’re going to give a little bit of details on that and we’re going to go broader, probably broader topics, talk about the impact to the overall. We’re still going to talk about companies, but we’re going to give broader impact. Spend a little bit more time on that. Give us some feedback. We’re always looking for feedback on the show. What do you like, what do you dislike? Thanks for tuning in. Take care. Have a great-