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:
- TSMC & Broadcom Eye Control Of Intel?
- Lenovo Q3FY25 Earnings
- What HP Acquiring Humane Means for AI
- Groking Grok-3
- Apple iPhone 16e And Its New Modem
- Microsoft Majorana 1 Quantum
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:
Daniel Newman: Hey everybody. Welcome back to another episode of The Six Five Podcast. It’s Friday hat day here on The Six Five. I wear hats all the time because I’m bald. But Pat, why are you sporting a hat? I mean, you’ve got a great head of hair, super handsome and thin. What’s going on over there?
Patrick Moorhead: Yeah, so I got to be honest with you, I was at 150 pound dumbbells and this morning I may have dropped one on my forehead.
Daniel Newman: Wait, I thought it was curls. You didn’t smack yourself while you were doing curls.
Patrick Moorhead: No, no. I do curls with 200.
Daniel Newman: I understand.
Patrick Moorhead: I just want get that out there. No, I actually, I want to give a warning to everybody when you’re taking your shirt off in the morning at 5:00 A.M. getting into your gym clothes and you got to watch that nail. So I got to be honest with you that I actually scratched myself with my nail. But awareness out there, I just want to increase awareness. The dangers of taking your shirts on and off.
Daniel Newman: Well, you do it a lot now with this sexy flexy thing you got going. You pass the mirror, you got to your shirt off, look and you’re probably doing it while texting and then you probably didn’t trim your coke nail. I mean, your pinky.
Patrick Moorhead: I don’t know. Should we tell people that we may or may not do conference calls when we’re both getting ready in the morning?
Daniel Newman: Sometimes. I mean, you got to use all your time wisely.
Patrick Moorhead: Yeah, have to. But hey, I want to give a shout-out to Lenovo for this awesome Ducati hat. They have a relationship, a strategic engagement brand and tech-
Daniel Newman: That MotoGP, or is that just Ducati? What is it?
Patrick Moorhead: I think it’s MotoGP. I’m not a big motorcycle guy, but I think it’s a pretty cool hat.
Daniel Newman: I like all things that go fast. Hey, you know what my hat’s for? People always ask me. They never know what the A is for.
Patrick Moorhead: Yeah, it’s the Oakland A’s, right?
Daniel Newman: No, no, no, it’s not. Don’t be a *censored*.
Patrick Moorhead: Well, tell them.
Daniel Newman: It’s A for A-hole because that’s what the average person thinks I am.
Patrick Moorhead: Can I get one? Can I get one?
Daniel Newman: This is for the Arsenal. This is the old classic Arsenal and if anyone knows me well you don’t even have to know me that well knowing that there’s three things in the world I care about tech, McLaren F1 racing and the Arsenal, of course, family, friends, that stuff. But that goes without saying. But yeah man, listen, it’s 250 this week. This is kind of like a five-year anniversary. It’s not exactly because 260 would be 52, but we missed weeks here and there probably no more than a couple a year. But dude, 250 episodes, I think we deserve some kind of award for doing that.
Patrick Moorhead: Not canceled award. What is it?
Daniel Newman: I don’t know. I don’t think we get canceled anymore. I think we’re in this weird period where all the stuff that got you canceled no longer gets you canceled. I read an article yesterday, it was kind of interesting, that it’s not cool to be progressive anymore. So apparently it’s cool to be cancelable or do things that were at one time cancelable over the last couple of years. I guess we’ll kind of see how that all works out. I’m just going to try to refrain from saying anything too stupid. I think we should both do that. Stick in our lane a little bit. I think there’s a lot of intersection with politics and macroeconomics and technology. I think these things are, but I think there’s a point where you’re talking about it within Foundry business and how that impacts the world versus us getting into an opinion on the war and the conflict in Ukraine. I just think there’s certain things that you know a lot about and there’s certain things that I don’t and I’m going to try to stick to the things I know a little bit more about.
We got a great show this week, Pat. A lot to cover. Not a crazy week. Next week by the way, we celebrate NVIDIA Day. I don’t know if that’s still a thing. I mean, I feel like the kind of enthusiasm is winding down, but the growth is still there. But this week there was a flurry of some stuff. The week started out with some big Intel news. We’re going to talk about that. Lenovo had their earnings this week. HP acquired Humane. What does that mean and should we care? Let’s grok a bit about Grok 3. The newest model came out. Some impressive numbers there. Apple came out with its new SE. I know it’s an E, but that’s the innovation of Apple these days. They took an S out of the name and I’m really cracking myself up this morning. And then Microsoft had a major breakthrough in matter with Majorana. They’re new. Are we calling them QPUs? I don’t know. It’s a new quantum processor. They’re saying, they didn’t call it quantum supremacy because we don’t say that anymore, but are we on our way to a million logical qubits here, Pat?
So, so much going on for everybody out there. We’re going to talk about some publicly traded companies. This is not investment advice. Pat, I don’t think there’s a topic that you and I grok about that gets more attention in the press, the media, in the technology circles and on X, than when we’re ripping on Intel. Let’s talk about what’s going on there. I mean, gosh, it feels like they’re getting their *** together, but now we’re hearing, I mean the Foundry is about to get spun off and you’re getting calls to get rid of their chairman. They want you to single-handedly make that happen somehow. And Intel is going to be what a division of Broadcom pretty soon. Is there any merit to this stuff?
Patrick Moorhead: Yeah, so Dan, if you remember about six months ago there were a flurry of, maybe it was three months ago, a flurry of rumors and you and I were weighing in and then they just kept coming, but so many of them were nonsense. So I just stopped reacting and then I felt that I had to get pulled in because of a lot of articles that came out. Bloomberg, Wall Street Journal, New York Times, all of these. And here’s the latest, and I’ll start kind of chronologically, that TSMC was going to buy 51%. Well, Frank Yeary, board of directors head for Intel, is just a story what was written. He’s been shopping with the administration for TSMC to buy 51% of Intel’s fab business. And then there was another one floating out there that, okay, well, Broadcom would buy it. And I saw two rumors they’re going to buy the whole thing and then there was, they’re just going to buy the product version of it otherwise called Intel Design.
So I was getting a bunch of calls, invited on shows to discuss that and essentially I just stick to my guns, which is any TSMC buying 51% of anything Intel is just complete and absolute nonsense. And it gets back to a couple facts, and dispute it with me on X if you’d like, but you can’t just buy a Foundry. It’s not just a factory with tools and people, right? There are two main assets. Maybe we can argue there’s three, which are the property, plant and equipment and then the transistor intellectual property. Giving a foreign company 51% ownership of the transistor IP is just not going to happen. It goes against everything that the current administration and the previous one and the one before that stuck to. And I probably got five or six models going in my mind.
And then it was funny, a couple of hours after one of these stories came out, the administration finally went out there and said, “Okay, we’re not interested in… “. Said, “Trump may not support foreign firm operating Intel’s US factories. White House Official says.” Knew this with that hilarious Bloomberg article and I’m glad the administration set that out there. And then there was an article out of The New York Times that I contributed to where they asked me, “Well, okay, if TSMC buys Foundry and Qualcomm buys design… “. But that’s another one. I can’t believe we even forgot about that one. That was in the Qualcomm article. “What would that make Intel?” And it’s pretty clear if both of its parts get sold off, it ceases to exist. But I think a lot of that is nonsense. Like I said, I do think a scenario that could work would be a minority investment.
Because here’s the deal TSMC is looking at 50 to 100% tariffs on their, and I think that is the level that would get the NVIDIAs, the Qualcomms, and anybody else who goes through TSMC would get their attention. So TSMC is eyeing the ways that we can make that go away. And the other thing is positional power. Intel 18A, right? I’ve been saying for a long time, and I think you have too, that hey, performance at this juncture of the process, performance looks good, power looks good, and this is for big die, using high performance transistors as opposed to mobile low leakage ones. It looks very good. And then boom, I think yesterday at a forum, Intel disclosed that even their SRAM from a die area perspective was actually the same as TSMC 2. And that was one of the criticisms that we had before was the SRAM size. So Intel has a lot of power. 18A is real. I will caveat that until I see tens of millions of 18A parts coming out from Intel, which by the way, 18A is a year ahead or so of TSMC 2. There’s still stuff that could go wrong.
Daniel Newman: Yeah, there’s a lot of good stuff in there, Pat. People aren’t really talking about this. But since the late-nineteen eighties, TSMC had over a hundred billion dollars of government investment made. Now again, that’s over three decades. This is not over a short period of time. And I think the amount of investment that would need to be to make Intel at parity would probably fall into a similar number, especially if they want to accelerate that. Now, Trump is all about American exceptionalism and American exceptionalism means that he does not want to be the President of the United States at the time in which the United States lost its technological leadership, because it basically lost the ability to produce leading edge silicon here in the US by US companies. So TSMC is a bit of this middle ground. TSMC is a partner. TSMC is a friendly to the US to US fabulous designers. And in fact, a lot of Intel’s woes are because TSMC has been so good to US fabulous designers that none of them want to use Intel’s foundries.
On the other hand… Hey John Fort. On the other hand, TSMC is not an American company. TSMC is in a country that is still debatably part of the China Empire. Now again, they don’t see, but China doesn’t see it that way. China’s a powerful nation building up its military. It’s shown with, again, DeepSeek is debatable exactly how it was done, but it’s shown what I would call a relentless pursuit and wherewithal to figure out ways to get what it needs to get to a parity. And get to competitive level it needs to get to make sure that they don’t lose the future on AI being held back by policy. So having said all that, I tweeted something last night and it was really pretty basic and it was that Intel would be idiotic to sell right now. I know it’s crazy, but Intel’s worst was like three years ago. I know nobody saw that because it actually takes time for a company that big-
Patrick Moorhead: Dude we saw, we all saw it and I got put, we both got put in the apple, but dude, I called the A&E thing early on.
Daniel Newman: The numbers didn’t show though because it took some time for the P&Ls to follow-
Patrick Moorhead: But that’s chips man, you know that. If you screw up architecture, it’ll take you minimum four years, you screw up transistor process and fab economics could be a freaking decade.
Daniel Newman: Yeah, no, no, I get it. So I’m just saying you and I both know from the work we do, right? Technological insights have a value of N and investor insights have a value of N times X times some exponential square. People just care more. And so the numbers have been so bad. This company’s been trading at close to book value, I mean at a literal deficit to book value for some period of time. And basically over this period of time, the last couple of years… And by the way with TSMC, the company actually dug itself out of a bit of a hole on client. It got itself to a period of time where it could build a competitive part, protect its market share and keep itself from basically flagellating underneath the weight of all that was being done by Qualcomm, AMD arm and others to where I’m not saying they’ve regained share, but they certainly have kind of ceased to shed further share.
And by the way, we did a huge report on AI PCs a feature, and we actually talked to 850 ITDMs. Intel’s still the highest priority preference part for AI PC, a lot of that’s legacy. Intel still sits with almost 60% of preference right now. And so anyways, Foundry, so let me just because we could spend most of the show on this topic. Broadcom is not going to buy Intel products. I’m just putting out there, I do not believe that rumor. First of all, Hock could do it. Hock could go in there and actually cut the weight and shed it. I just don’t see that he needs it. I really don’t see that he needs it unless this is some sort of forced marriage and makeup play that Trump does because of the Qualcomm history that he does a favor of Broadcom and sort of wants to prop up Intel.
But the Intel products business separated from the foundries actually a good business on paper.
So if you’ve actually cut that out and took it out public, it creates a lot of adjusted EBITDA and it would have good EPS. The Foundry is another story, but the Foundry in my opinion, to maintain American exceptionalism requires that a hundred billion of investment by the US to make sure that in no way does the US fall behind. Now that means the US is going to be subsidizing it. That means the US is going to have a say in it. But I actually think to some extent the US needs to reapply the play that we made to help Taiwan. Because the US made Taiwan, let’s be very clear that Taiwan invested a hundred billion dollars, but we sent all our best designs and our best companies to Taiwan to manufacture their chips.
We could easily implement policy to turn that around through tariffs and through incentives to basically bring that back and say you make your chips here in the US and 18A is a good process. We don’t know, like you said, we don’t know everything. We haven’t seen volume, we haven’t seen scale. But all the early readouts suggest that Intel and TSMC could be much closer to parity over the next year or two than they’ve been in a long, long time. And if the US wants to lead the world in AI, they need to lead the world in Foundry. That’s my take, that’s why I think it’d be idiotic to sell. That’s why I don’t understand what Frank you’re here is doing here, but I do understand that in some ways returning value to shareholder in the near term doesn’t always make sense over what is best for long-term economics.
All right, let’s move on to something that’s not probably as polarizing among our community, but something that did take place yesterday. A company that sells a lot of Intel, by the way, Lenovo had its earnings. Company that’s another one of the sort of bellwethers for infrastructure for PC and client and they showed really good numbers, Pat. Lenovo has had some ups and downs its devices. Business has done pretty well consistently been sort of a core stabilizer of that company. But the company under the leadership has really been trying to pivot to the infrastructure side, winning more of the middle of the market, not just the hyperscalers where they’ve had a lot of success on compute, they’re trying to add a lot of services. They’re trying to really be a bit of a hybrid player in terms of hybrid AI and all the way up and down and Pat, the numbers showed 20% revenue growth this year to 18.8 billion. They doubled their net income, which the market should really like. And on top of that, their infrastructure group. Now again, this is the GPU boom, 59% growth year-over-year.
So clearly Lenovo has found a path into the GPU space. They’ve got some really great liquid cooling technology that lets warm water cooling. I think that’s probably playing very well in their favor. But saw growth across the board, Pat. 12% in devices, I think it was 12% in services, it was 59% in infrastructure. And their non-PC revenue mix has gotten close to half the business now, which has been a big diversification play for the company. So good quarter for Lenovo, we talked to one of their presidents, Ken Wong, you and I last night late. And my overall perception is the company is doing well. I always worry a little bit about the dependence right now with the NVIDIA boom, but it’s also very wise, whether it’s Dell, HPE or Lenovo, you’ve got to run with these things when they come. So Pat?
Patrick Moorhead: I just want to say this again, 20% year-on-year revenue increase. I mean that is absolutely stunning to me. And while the infrastructure group got a little bit of a slow start in AI, these are the results that they have finally clicked in. Lenovo, particularly with hyperscalers, had some designs that their competitors were earlier with, but now they are full-blown in there and they do have a very important technology called Neptune Liquid Cooling that is distinctive out there. Dan, you and I are going to have the chance to chat with Ashley next week, I believe, who is the new leader of their infrastructure group. And listen, with a 59% increase in revenue, your profitability should turn around if nothing else across some of your fixed expenses. Profit margins are light. And to me that is evident of concentration with hyperscalers where you’re really running an ODM model and they call it the ODM Plus model out there, but a lot of cash coming through the door, which is good.
SSG, I think it’s almost a carbon copy. They’re really impressed with when they’ve been able to do on the software and solutions side. Daniel, our firm’s track the Accenture’s of the world, the Deloitte’s of the world, folks like that, and a very similar level of sophistication coming out of them. Now they’re in IBM consulting. They’re a much smaller entity than any of the above. But it is interesting how you would expect them to know more about the technology than these other, the base technologies, the infrastructure, the PC. But I think SSG is one of the most interesting parts of the overall Lenovo story because it really is talking to the CIO as opposed to just the purchasing agent here.
I am looking forward to see if SSG can help the infrastructure group gain share inside of the enterprise market. When I look at Lenovo, the data center group, they’re really strong with the hyperscalers and they’re really strong with the small and medium business. They do very well with HPC, with labs and things like that. So I’m interesting to see if SSG can give Lenovo Enterprise overall lift in its ISG group. One final point if you missed it, there was a $2 billion investment that Lenovo made in a partnership in the Middle East with a company called Alat. You may have seen announcements from a who’s who in tech of Saudi and Qatar, Qualcomm did some big announcements over there as well, but this is another example of Lenovo’s multinational strength.
Daniel Newman: Yeah, there’s a lot going on. I think it was Leap a week ago. There was quite a bit of big news that came out of the Middle East, definitely.
Patrick Moorhead: What are we going to take The Six Five over to the Middle East?
Daniel Newman: I don’t know. We should talk to our production and sales and business teams and say, “When are you going to take us over to the Middle East?” I’m down to go. I hear great things about Dubai. So that would probably be my first choice, but we could probably go to Riyadh too. Another nice place. Or of course Tel Aviv is another interesting and exciting place in the Middle East. All right, Pat, is it humane or inhumane for Humane to be humanely shut down, sold. Was the putting down of Humane done in a humane way or what happened there, Pat? HP takes over the pin.
Patrick Moorhead: Man, that is so good. I don’t even know where to go from there. I should just-
Daniel Newman: World’s best moderator.
Patrick Moorhead: No, totally. You should just do the mic drop right there. So a little background. Humane was a software and device company that did not have a good experience out there. Essentially it was a pin using generative AI to do different things and did not go well for them. HP has acquired the Humane intellectual property and also the software called CosmOS, and it was a bit of an aqua hire and $116 million investment based on what was a $240 million valuation. So why on earth are they doing this? Well, it’s all about CosmOS and that’s C-O-M big O-S platform. And if you think about a layer and arbitration and management layer for agents across all of your devices, PCs, printers, conference room systems, all the consumer gadgets. And then you extend that with the different modalities of AI, which is on device, cross device and cloud, you can start to see the magic. And while HP didn’t use the word agents, if you go and do research on Cosmos and Humane, it was all about agents.
So you can imagine automating workflows between these devices, handling different requests on prem between different users, maybe without going to the cloud, which I think would be would be super interesting. There’s a lot of work to be done here. I know there’s a lot of talk about, “Oh, HP acquires companies, doesn’t work out.” A lot of that was in the past folks when it was literally called HP and it was a software and enterprise and PC company. That was a long time ago folks. So let’s give this a shot. It is a challenging environment and I think there’s also, if I may say that this might be an indicator that Microsoft isn’t moving quickly enough on hybrid AI across devices or HP sees this as the ability to get ahead of it and do things before Microsoft pulls this all together. So strategically I get it, 100% on where this is coming from.
Daniel Newman: Yeah, Pat, I think this is mostly an intellectual property. Some capabilities. I think there’s this ambient AI opportunity. We’re sort of moving between our devices and our handsets, our PCs now that are AI driven. There’s been a bit of this sort of meta and AR VR XR thing that’s come in and out of conscience of our consciences, but in the end, we have not landed on a platform. Like I said, it was interesting when I got pre-read in on this and I was reading, I was like, hmm, the pin didn’t work because do I really need to pin AI onto me when I’ve got this thing on me all the time? With a more powerful compute engine you can wear in your ears, you can put glasses in your eyes with audio that’s limited for you to do your readout. So I’m trying to figure out the play.
And so to me, I think some of the play is about what’s listening, meaning does HP want to maybe create these kind of edge devices or this edge sort of AI experience on a device that’s private. That can kind of move with you all the time and be your sort of ongoing agent assistant without necessarily giving data here on the device? I’m not sure. I think there’s some interesting engineering and I mean at the price they paid, it feels like they were mostly acquiring a team. I think it was like 115 million, which is crazy. It’s a big number. But in the era of unicorns and billion dollar businesses and companies, I was sort of trying to understand the play. I think AI in, it’s kind of like, Pat, you do a lot of health stuff. It’s like your phone could in your watch, could do most of it, but you wear these additional accessories because they do something that’s more granular to your need than what you can get on the kind of off the shelf. I mean, you can give me the read and if that’s true.
My point is that’s kind of what I see here is there’s this third AI layer, there’s the PC, there’s the phone, and then there’s this kind of pervasive ambient experience that sits in between them. I think there’s a lot of work to be done to message that. I think the average person’s going to look at their PC and phone as enough. And so I think that’s the real opportunity here is can HP help the market understand this sort of third leg of the stool where ambient pervasive connectivity to maybe a device that’s not connected to everything else that can be used for this makes sense. Adds value, provides data protection, sovereignty, security, but also gives those agentic and assistant values, Pat. I’m still kind of trying to read through this one, but that’s kind of where I landed.
Patrick Moorhead: Good comments on that. Yeah, the whole ambient computing thing, if you look at the acquisitions that HP has made, whether it’s Poly, all the gaming accessories, they have printers and they’re pulling all these things together. Definitely, I mean, it makes a lot of sense for me based on that and then what these devices have to do.
Daniel Newman: Yeah, no, absolutely. So that’s kind of where I stand there and I mean we’ve seen it a little bit with Alexa and the home pods and all the things. Is there a work story here? Those things have never quite made it in the workplace, but it always kind of felt like those could be great ways to add intelligence and smarts into every room, every space. This is kind of a personal version potentially of that I guess.
Patrick Moorhead: Interesting.
Daniel Newman: Let’s grok. You want to grok a bit?
Patrick Moorhead: Let’s grok baby.
Daniel Newman: So this week on a very short timeline, Elon Musk announced that he’s releasing Grok 3. It was announced a day before and the next day there was a live stream. It came out. Grok 3 is basically the most advanced features are available for you to try on the browser in the web, and then they’re building the features into X. And then of course you can do the app, but the app is he said so eloquently, it takes about a week or so for the new updates to be applied and approved if you’re trying to use the Grok app on your phone or on Android or Apple. So you’ll have to wait. But out the gate Pat, we shared some stuff. I mean there were some various shares that came out about Grok 3’s performance and it was very good. Now having said that, there was a little bit of an N minus 1 comparison going on that I had to look at both because with the N minus 1, Grok kind of won everywhere.
It won in math, it won in science, it won in coding. But then I did see some of the more, some of the pundits out there that aren’t named Pat and Dan shared a second version of it. It kind of incorporated reasoning test time, compute comparisons, and included o3, which is the largest late model from OpenAI. And in those categories, Grok 3’s reasoning and mini reasoning still both performed incredibly well, but did come just behind. But this still brings a really interesting thing about what’s going on with reasoning, with test time, with efficiency. I mean, XAI has not scrimped on investment. I think it’s what, 200,000 GPUs now, and I think Musk talked about adding to that this week.
So by the way, I think NVIDIA is going to be okay from what I’m hearing here, but it is interesting how quickly XAI we’re seeing it with Llama, we’re seeing it with Gemini. There does feel like there’s a lot of pressure on OpenAI at the front of the pack where I increasingly, as I see SoftBank suggesting a $250 billion valuation. Where it seems that others are able to build interestingly and competitively close products to do the key functionalities, like I said, the math numbers, the science numbers and coding, which actually Grok apparently outperformed the OpenAI models. That it’s moving really, really quickly, Pat. So I think it’s a really interesting time in the models and rather than just rambling on about this one.
I feel like this is an increasingly commoditized part of the market. I feel like how we apply the data to them, but Pat, I don’t know, over the last week I’ve played with deep research, I played with OpenAI Deep Research, Perplexity, I’ve been playing with Grok 3. I mean the speed of innovation in which taking place here is absolutely mind-blowing. And Musk is moving very, very quickly. He’s really putting the screws on and turning on Sam. And man, by the way, just a side note, the debate between those two, never getting rid of Twitter, never getting rid of X, man. I just absolutely find it entertaining.
Patrick Moorhead: It’s intoxicating, man. I doom scroll way too much on it. Hey, on Grok 3. I think you eloquently laid out the case for yeah, it’s really good at math and science and coding. It may not be as good as o3 standard, which came out in December, but just the fact that it’s even in the hunt I find absolutely miraculous. Now, one thing we do know is it took 200,000 GPUs, which may or may not indicate that scaling laws are in effect. I think the latest Llama 3 took 50,000 GPUs. So if we compare that to 200,000, how does that scale? Is it a four x improvement? No, we don’t know. For o3 as an example, we actually don’t know how many GPUs they used. Only OpenAI and Microsoft would know that. So it’s kind of an unknown. I think what we should know though is the time it took to get this thing out was absolutely incredible. And that could say, okay, maybe the scaling law, if you put in time to market and time to pre-train and train, that’s what you should be looking at.
So I’m not there yet that there’s a direct scaling between how many GPUs and the capability. Could be the benchmarks aren’t up-to-date. Final thought is that Elon was very clear that he’s going to be improving this. It’s not going to be okay, we’ve smashed out Grok 3 and we’re not going to improve it and just move on to Grok 4. Kind of like we’ve seen with some previous model makers who were pretty much focused on the next best thing. I’m wondering if that could indicate that this came out quicker than possible or quicker than they wanted to.
But listen, anytime you can run up the score, at least based on these tests on o3 mini, o1, DeepSeek and Gemini 2 flash thinking you’ve done a really good job. The other thing we don’t know is how many employees has it taken to move this forward? So anyways, a lot of knowns, a lot of unknowns, but congratulations to the Grok team. The way congratulations also to some of the suppliers in there, like Dell for instance. VAST Data also is in there, and I’m sure I’m forgetting some folks. The head note is probably Intel, but I digress.
Daniel Newman: Yeah, absolutely. So we’ve got a couple more topics and a couple more minutes, so let’s keep running, Pat. I think both of us did quite a bit of media coverage, social and content on this topic of the new iPhone. Some of it was about the phone itself. Snore, hold on, ready? But some of it was about a new modem that has maybe changed a paradigm that has been difficult to change for a long time. What’s going on there?
Patrick Moorhead: Yeah, so listen, I mean, Apple has demonstrated that it can bring out subpar products and they’re going to sell great, particularly in the United States and certain countries in the EU, but that’s not the case for everywhere. And side editorial point, it will catch up with Apple at some point. They do have such a stranglehold and a walled garden for apps and peripherals that it really makes it difficult for anybody to get meaningful share. Out the other side of the mouth, I’ll remind everybody that the number one share in the entire world is Samsung and smartphones not Apple. So opening price point, OPP is 599 versus OPP of the current SE at 429. That’s a 40% increase in price there. There are a lot of things that would make it more costly like storage the display, the A18. Does have Apple Intelligence for those people who care today, which are very few but could be in the future, there’s some arguments out there that, “Oh, well the iPhone 14 is actually the new SE.? Well, you know what? Just say it and let’s Apple commit to that and let’s see a statement from Apple that commits an ongoing supply of the iPhone 14 at that price for x amount of years. But this is classic Apple not saying anything out loud. It’s, hey, people have to figure this out.
As you said, Daniel, an interesting point here. This included a discrete C1 Apple, I’m going to call it chip modem, chip sets. It’s not just one chip. A lot of talk about it being lower power, which I guess my response is, well, okay, give me access to the ability to hit that modem and let’s do a side-by-side compare. I’m skeptical Apple should open that up to folks. The other thing is that there’s no third party tests available for reliability, right? Think dead spots, think drops, think the inability to potentially get out that TikTok post and it’ll probably be slower. I have absolutely no evidence for that. But what I do know is there are no free lunches. It took Apple five years and a multi-billion dollar investment not only in the core IP.
And by the way, five years ago, Intel had a non-millimeter wave modem, which by the way, this is, that was available that Apple was doubling down on that they were going to ship with until they decided that they weren’t. And the other thing is it’s not integrated into the SOC. It’s like how really does this compare to an integrated Qualcomm modem on performance, on power, on reliability? I don’t have to see any tests to say that this modem, I’ll give you a 95% degree of confidence. It’ll be less reliable. It will be slower. And if we had the ability to actually hit the bus that the modem is on is going to be a lot higher than integrated modem in let’s say a Samsung device where the modem is actually integrated onto the SOC.
Daniel Newman: I’ll just touch quickly. I agree with you on your modem points. I think this has been a long time coming. Apple had to get this out to prove that they could break the chains of Qualcomm. Qualcomm’s modeled for this for a long time, so it’s a slower erosion than expected and it’s actually been a better outcome than the investment thesis that had come forward over the last couple of years. I still don’t necessarily believe it’s going to work well and find its way into the premium devices. I think this is kind of like let’s give the crappy stuff to our low-end customers and see if they notice. I mean, look, the SE phone suck. Anyone that had an SE phone, they were glitchy. I thought they were terrible.
I’m not saying that, it’s like something I give my eight-year-old. It’s never been a device that I would use every single day if I was going to spend in that tier. There’s so many better options from Chinese companies, from Samsung and of course, but Apple knows that it’s got this audience, it’s got this buying group. I was on Fox Business the other day and they asked me about it. I’m like, “It really doesn’t matter if they’re innovative. People still buy their stuff.” I’m like, “What’s going to happen with the stock?” I’m like, “Well, I don’t really do that.” But I’m like, “Let’s just say it’ll probably go up.” “Well, why is it going to go up?” Because it doesn’t seem to matter if Apple makes good stuff, people just buy it anyway. And it’s like the intelligence, it’s absolutely garbage. So part of this is getting a cheap device out that has Apple Intelligence on it to start exposing.
Now, what I will say is Apple has the grace of basically being able to get to AI when it wants to. I think you tweeted something the other day about how generationally bad their generative AI is, how you can’t believe, I think we all thought it would be, we’re all kind like, okay, it’s going to take them longer, but we didn’t think it would still be this bad two years after the ChatGPT moment, that it’s like they still have nothing of meaning. And I’ve turned Apple Intelligence back on because I was really hoping something had changed. It’s just not very good. So the good news is you get more users, more exposed, you bring more down and I don’t really think the quality of this device is good enough to do any real share shift from the better product. So this is all about kind of added share potentially for them I think. But if it is actually too good, Pat, then it’s going to cannibalize a higher margin better product, which is also dumb in my opinion. So-
Patrick Moorhead: Not at that price, man. Not at that price. It’s just weird. And I’m wondering if they’re holding some tariff dollars kind of just in case there.
Daniel Newman: I got a great delivery. I got three red eggs, man, a little piece of sourdough here.
Patrick Moorhead: Gosh.
Daniel Newman: Some blueberries. Just came out my desk. I got to get done with this pod so I can eat this stuff. All right, let’s get to the last topic. Quantum, Pat, I feel like sometimes you got to be a physicist, a PhD, a researcher to really talk quantum. But I think what you and I try to do for the world is make it make sense. So this week we saw major investments from Microsoft with quantum, Majorana. Is that how you say it? Majorana? It’s a new kind of topological core architecture and they call it a topoconductor. So it’s not a superconductor, it’s not an ion trapping, it’s not an atom splitting. This is a topoconductor, which they’re calling a novel type of material that can basically control Majorana particles. I’m reading this.
I’m being honest, this stuff is super technical, to produce more reliable and scalable qubits. What most people don’t understand is to really get to high value quantum, there’s numbers thrown out there. You need to get to a million logical qubits. I’ve heard 100,000, I’ve heard 10,000, but it’s a lot more qubits than we’ve been able to stably create with any of these different technologies. A lot of cases you’re hearing about continuum’s at 32 or 64. We’ve heard IBM talk about numbers getting up into the thousands now, but getting to a million qubits so that we can actually start to use quantum with classical to solve some of these really complicated problems seems to be the goal.
So you hear about things like Willow and it’s going to break encryption and Bitcoin’s going to get broken. And then you see things like this. And what I did come away from this, Pat, is really come to the conclusion that I do think the big tech companies will be the winners here. I know we’ve got all these little kind of meme stocks, the Regettys and D-Waves and IONQs that are all sort of, and IONQ’s been the most interesting. They’ve won a lot of business and contracts from government, certainly has been able to pull away from kind the other meme quantum names, the beverage companies. But I think what we’re seeing is that the big tech companies see this as the next frontier. Satya called it the transistor moment. Him and Elon had a little back and forth. I don’t know how true that is. My sort of quantum experts in the background gave me some advice that there’s a lot of claims here that require proof that the scientific community is kind of kicking around whether this is or isn’t what it is.
But the bottom line is smaller, faster, more greater coherence to reach more logical qubits. You talked about the ability to get a million qubits on a chip that could fit in the palm of one’s hand. That is breakthrough. And so I’ll leave this thought here Pat, on the same day Apple announced a tuned down iPhone with a tuned down modem that’s not part of the SOC for $600 that offers features. And meanwhile Microsoft delivered a million qubits on a chip potentially. So I’ll take that every single day. I’m not sure it’ll get there. But Pat, this was a really big breakthrough moment, certainly in the research community.
Patrick Moorhead: Yeah, I want to give a caveat here. I have a quantum analyst named Paul Smith-Goodson that is about a thousand times smarter than me and it is very hard to even put a comparison to what other companies are doing. But this is a breakthrough and in every sense of the word as we have a brand new architecture, it’s a topological course. So that in itself and what gives this credibility, DARPA chose Microsoft as one of two companies. I think PsiQuantum might have been the other one. And what this award was was essentially, “Hey, who has the biggest risk return tradeoff?” And the goal with DARPA was to have something scaled utility scale, quantum computers by 2033. That just gives you a sense of where DARPA was and Microsoft and PsiQuantum, I call it like throwing the long ball. If you want to go big and you want to go big in 2033, these are the two companies that you would go with.
There are companies who are ahead of Microsoft, namely IBM and even INQ and even Continuum on a nearer term type of deliverable here. But again, I don’t want any of that to take away from all that Microsoft did, but at least Paul is telling me that Microsoft is behind in terms of years. But it can make up ground and pass the other technologies if Majorana is as reliable as it is expected to be. But again, this is still in the area of research. We are not too productization at this moment and I think we need to keep all winner take all exuberation of anything with a grain of salt because there is a lot more work to do, a lot more research to do. And just when we thought that, okay, we can put everybody into a nice mold, whether it’s trapped ion superconducting, and I’m sure there’s others, we’ve got this topological Majorana out there that surprises all of us.
Daniel Newman: And there you have it everybody. I am four minutes late for corporate BS and I have to run and Pat, I’m sure you’ve got a day chock-full. It is the weekend. I will be hitting the weights hard. I’m going to try to deadlift so much weight. I pass out on the way up just in case. And then I throw up on the way down.
Patrick Moorhead: It’s a video, send me video buddy.
Daniel Newman: In all serious, I want to just thank everybody out there for tuning in, sticking with us. Every week’s going to be a big week. We got more to cover. We’ll see you all very, very soon. Bye-bye.
Patrick Moorhead: Bye-bye.
Author Information
Daniel is the CEO of The Futurum Group. Living his life at the intersection of people and technology, Daniel works with the world’s largest technology brands exploring Digital Transformation and how it is influencing the enterprise.
From the leading edge of AI to global technology policy, Daniel makes the connections between business, people and tech that are required for companies to benefit most from their technology investments. Daniel is a top 5 globally ranked industry analyst and his ideas are regularly cited or shared in television appearances by CNBC, Bloomberg, Wall Street Journal and hundreds of other sites around the world.
A 7x Best-Selling Author including his most recent book “Human/Machine.” Daniel is also a Forbes and MarketWatch (Dow Jones) contributor.
An MBA and Former Graduate Adjunct Faculty, Daniel is an Austin Texas transplant after 40 years in Chicago. His speaking takes him around the world each year as he shares his vision of the role technology will play in our future.
