NVIDIA Comes for Windows PCs - DTNS 5280

0:00This episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi-agentic AI. They already deployed one. It's called Chat Concierge and it's simplifying car shopping. Using self-reflection and layered reasoning with live API checks, it doesn't just help buyers find a car they love. It helps schedule a test drive, get pre-approved for financing, and estimate trade-in value. Advanced, intuitive, and deployed. That's how they stack. That's technology at Capital One. When you're a maintenance engineer in a beverage manufacturing plant, you keep production lines moving and quality on track because there is no room for slowdowns.

0:39With Grainger's vast selection of high-quality motors, sensors, belts, and hard-to-find parts, you can get what you need fast and all in one place so nothing gets in the way of getting the job done. Call 1-800-GRAINGER, click grainger.com, or just stop by. Grainger, for the ones who get it done. Still waiting in line? Again? That's time you will never get back. Save time and money with Stamps.com. Over 4 million businesses have skipped the line with Stamps.com. Join them to save up to 90 % off carrier rates from your computer or phone right now. Print postage for certified mail, registered mail, and packages in seconds.

1:20Then schedule a pickup right from your home or office. For a limited time, go to stamps.com and use code podcast for a free welcome gift. Taxes and fees apply. I'll be honest. Nothing tests my patience quite like sitting on hold, listening to the same four bars of music on repeat. Turns out, all that time adds up. The average person spends over 40 days listening to that dreaded hold music. 40 days. This episode is sponsored by Parloa, the AI platform built to make that a thing of the past. Parloa's AI agents are available 24 hours a day, 7 days a week, fluent in any language, and smart enough to remember every customer across every interaction, because no one should have to explain themselves twice.

2:03From payment processing to roadside assistance, flight rebookings to appointment scheduling, Parloa delivers the kind of customer experience that turns a one-time buyer into a customer for life. See Parloa's AI agents in action at parloa.com. That's P-A-R-L-O-A dot com.

2:29this is the daily tech news for monday june 1st 2026 we tell you what you need to know give you important context and try to help each other understand today nvidia is coming for the windows pc and everybody is gunning for the macbook neo i'm tom merit and i'm rob dunwood let's start with what you need to know with that big, big story.

2:53Yeah, so we got Computex kicking off in Taiwan. Lots of chip news, lots of hardware news. And alongside Computex comes GTC Taipei, the GPU technology conference from NVIDIA, which kicked off with a keynote from CEO Jensen Huang. Kind of has a dual role, honestly, as kicking off both Computex and GTC Type-A. They run concurrently. The two main announcements from NVIDIA were the next generation Vera Rubin platform for data centers is in full production now. All those data centers that want to buy these chips and put them in can now get them. The other big announcement we'll get to in a second is NVIDIA is making a CPU for Windows PCs.

3:37Not the first CPU they've made. They've even made a CPU for a desktop in a workstation, but this is for consumers. Now, let's get to Vera Rubin real quick. Wong says that platform will have the lowest token cost in the world for inferencing. Inferencing is the process of answering queries from an LLM. And announced that Anthropic, OpenAI, SpaceX, 4XAI, Dell, CoreWeave, and Oracle, all on board, all want to order Vera Rubin gear, and it's going to start shipping to them soon. Okay, so if you're into the data centers, that's good to know. You know that that stuff is going to be available. It's going to go in the data centers.

4:13It should be more efficient. That's going to help inferencing down the road. The RTX Spark is the more unexpected announcement. That is the name of a 20-core Grace CPU developed along with MediaTek, which includes 6 ,144 cores of Blackwell GPU performance. They say it's about one petaflop if you're judging your AI performance, and has unified memory up to 128 gigabytes of lp ddr 5x memory unified means it can take advantage of the whole thing that's one of the things that makes apple silicon uh so special uh but it also maxes out at 80 watts so you have less power than an equivalent rtx 50 you have slower memory though it is unified uh so it's it is different than an rtx 50 otherwise it's blackwell so it's the same as the RTX 50 series of GPUs.

5:08RTX Spark CPUs will come in laptops, desktops, and the DGX station. That is the next version of that workstation I mentioned earlier. NVIDIA promises that will deliver data center grade performance in a desk side form factor, since it's too big to put on top of the desk in most cases, or you wouldn't want to take it up all that space, desk side. In fact, RTX Spark seems to be a rebrand of the DGX Spark. That was the chip that launched in the first DGX Spark workstation. And being Blackwell-based means it could take advantage of DLSS upscaling and frame generation, just like RTX 50 GPUs. Now, Huang said the first models would be targeted at creators, AI developers, and gamers looking for very thin and light laptops, slim laptops, portable laptops, or compact desktops.

6:02And when he said gamers, he meant it. NVIDIA wants to overcome the barriers to gaming that a lot of ARM-based PCs have faced. In fact, they told Ars Technica that they are working with Riot Games to support League of Legends and Valorant. They're working with Crafton to support PUBG. And they're also working with the developers of Easy Anti-Cheat, BattleEye, and Denuvo. Anti-cheat has been one of the big barriers to gaming on ARM in some situations. So they want to overcome all of that. NVIDIA also says it's working with Adobe to rebuild the core of Photoshop and optimize Premiere. That's for the creators.

6:39And if you're wondering, TSMC, of course, is making these chips on their 3-nanometer process. NVIDIA says more than 30 laptops will be on the way, 10 desktops on the way. Asus, Dell, HP, Lenovo, Microsoft, MSI, Acer, and Gigabyte all going to make machines using RTX Spark and NVIDIA promises those laptops will be around 14 millimeters thick. So nice and thin. Microsoft in particular gave us some details on theirs. A 15-inch Surface Ultra laptop. It has a mini LED screen that peaks at 2 ,000 nits of brightness. The largest trackpad Microsoft has made and lots of ports. If you like USB-A, USB-C, HDMI, it's going to get all of those as well as a full-sized card reader and all of that be less than four and a half pounds.

7:25NVIDIA is working to make sure Windows can be more agentic on the platform and Microsoft, of course, is going to give us more details right away when Microsoft Build kicks off on Tuesday. We have a few more NVIDIA bits to get to, Rob, but let's pause there. What do you think of the PC chip approach? So there's a lot to talk about. The first thing I want to know, particularly with the RTX Spark, is because this is a desk side type of server. Does it come in beige to remind me? Oh, you mean the DGX? Yeah, yeah. The DGX station. Yeah, it must. It must come in beige. I think that's the default. It's got to come in beige.

8:03I'm just thinking of back in the day when you had the big giant computer that it was too big to be on a desk and it had to sit next to the desk. Those things are always beige and the size of a small dog or a medium sized dog. So they were quite large. Seriously, though, this is there's a lot here. The thing that kind of stands out to me is that NVIDIA may be jumping into the PC space as far as just creating chips for PCs. That that is a big thing, especially for someone like me who lives next to a giant hole in the ground called Intel. that they are still building, but just not nearly as quickly as they did once before.

8:40So the question is, can they really make a dent into what not only what the other ARM-based processors are doing with PCs, but can they encroach on what Intel is doing? That is very problematic from an Intel standpoint, because NVIDIA is just the biggest company in the world right now. So that is the thing that is probably of most interest to me, is that How good are these PCs going to be running hardware that we haven't seen them run before? Yeah, it's interesting to see how Windows on ARM has gotten so much better. In fact, if you remember the old Windows RT, which was the first attempt to run Windows on ARM and it did not succeed, that was on an NVIDIA Tegra processor.

9:25So NVIDIA was there in the dark, dark days of running Windows on ARM. Windows on ARM has gotten so much better. And NVIDIA is like, let's bring it over the line. Let's work with Adobe. Let's work with game manufacturers. Let's make it optimized in an attempt to take market share away from Qualcomm, mostly. I mean, yeah, they're going to take it away from Apple a little bit, but that's a bit of a different situation. But Windows to Windows, Qualcomm wants to take that share from Intel that you were just talking about. NVIDIA would like to take that share from both Qualcomm and Intel. So I think the other interesting thing is the knee-jerk reaction has been, well, NVIDIA doesn't care about consumers anymore.

10:05They're all about data centers. NVIDIA cares about any market where they can increase their sales. And this is one where they think they can increase their sales. And if it's consumers, great. They'll try to do that. They are a company and companies tend to like to dominate markets. If they can dominate another market, why not? We're building processors. Can we stick them in these things that people like by that they sit on their desk on the side of their desk that they carry in bags around with them if the answer to that is yes then nvidia by all means jump into that category so it it's it's an interesting thing i don't want to say that we're in the wild wild west of cpus because there were some things in the in the mid 90s to the late 90s that just kind of dwarf what we're happening you know happening now because a lot of this is still ai based but it is it's very interesting conversation that we're having when there are so many different manufacturers that are building processors that are going inside of hardware that we use.

11:00It used to be Intel and AMD. And that's all you really heard until you got to the big giant servers that were running in data centers. Now you're hearing a lot of different players in the space. And although the flavors of Windows may be a little different, you know, ARM-based as compared to, you know, the Intel, the x86, you know, architecture and, you know, what has come after that. It is very interesting conversation. There's just so many players that are getting into the PC making space or the laptop making space or the game desktop making space. This is the most competitive CPUs have been since the 90s, for sure.

11:35And it's good for consumers because you're going to get a lot of choices. The one thing NVIDIA, I think, is going to run into, and they've been a little cagey about the prices on this because they're going to be high. They're going after creators. They're going after developers. They're going after gamers. Those are sectors of the market that are used to paying a premium to get powerful, efficient, portable abilities to do their thing. These are not for the MacBook Neo crowd, right? There are others coming for that. We're going to talk about those a little later in the show, but this is at the high end.

12:09But as a consumer, you've got so much choice. You want that kind of thing? You've got multiple options, not just NVIDIA getting into this race now. You've got more affordable options. We're going to talk a little bit about some creative options in bringing more affordable parts to PC builders, whether you'll like them or not, we'll see. But yeah, this has become the most competitive landscape in recent memory. Here's a question for you, Tom. Is running LLMs, is running AI on a desktop where you're not in a cloud, you're running it on your PC, is that really appealing to a lot of folks from what you're hearing?

12:51Yeah, I feel like that's why they segment the DGX station away from the RTX Spark. They're like, look, if you're a creator and you want to do some agentic stuff, we're all about that. We're going to help Microsoft optimize agentic because we know a lot of you want to do that. You want to have agents that take care of things for you. If, however, you're doing heavy coding, you're doing something that is a lot more process intensive then we're going to have that desk side DGX station for you. But I think to kind of get back around to answering your question, when they're talking about agentic stuff, yes, NVIDIA is banking on you will want to use agents more often, whether they are calling the cloud, you know, or not.

13:36I don't know if it's on device. If it's on device, though, that's what the DGX station is for, is like you can do everything on device with that one. Interesting times. Yeah. Yeah, indeed. A few more bits here, also pointing to the fact that NVIDIA doesn't want to have all its eggs in the data center basket. NVIDIA working with China's Unitree to mass produce humanoid robots that work out of the box. A lot of the setup with humanoid robots has been a barrier for entry because you have to do a lot of customizing, a lot of training, a lot of tailoring it to the situation you want it in. uh nvidia wants to make that so that you just open it up turn on the robot and tell it what to do and it goes and does it uh the company also released the isaac groot reference design that's gr00t uh humanoid robot platform uh that reference design platform is meant to be used with a unitree h2 plus humanoid robot sharpas five-fingered hands and the nvidia jetson thor your onboard compute.

14:38You are going to see a lot of people use this because it's a reference design. You have to buy the stuff yourself. But AI2, ETH Zurich, Stanford Robotics Center, and UC San Diego all saying they're going to use that open platform. And speaking of open platforms, NVIDIA released Cosmos 3, an open world model to help all kinds of physical AI, not just robots, but autonomous vehicles as well, understand and predict real world environments. So it models how the machines move, not just how the world looks around it. Partners for that already include Agile Robots, Black Forest Labs, and Runway. Robots and AI.

15:18That's just, that's just, that's where the, I can't say the internet. That's just where the industry is. It's like - That's where tech is right now. Tech right now is about artificial intelligence and robots. So you know, I've got my fears about Why are we so hell-bent on building Skynet? But, hey, you know what? If it's good for you, it's good for me. Yeah, we've got Skynet. We just need the Terminator robots now. Well, folks, DTNS is made possible by you, the listener, and we have a lot of folks to thank today. So, new patrons, Annie, i.73, James, Wojo, Sterling, Huehenji, Callie, Andrew, Kenneth, Dennis, Renee, and Gray Mora.

15:59Thank you all. Yeah, I love Mondays.

16:28schedule a test drive, get pre-approved for financing, and estimate trade in value. Advanced, intuitive, and deployed. That's how they stack. That's technology at Capital One. Still waiting in line? Again? That's time you will never get back. Save time and money with Stamps.com. Over 4 million businesses have skipped the line with Stamps.com. Join them to save up to 90 % off carrier rates from your computer or phone right now. Print postage for certified mail, registered mail, and packages in seconds. Then schedule a pickup right from your home or office. For a limited time, go to stamps.com and use code podcast for a free welcome gift.

17:07Taxes and fees apply. Ever been at the checkout wishing you had more cash to spend? You're watching the total climb, doing mental math, getting ready to put your favorite yogurt back. It's the worst. Clio lets you access an up to$250 cash advance fast so you can cover your groceries without the panic. No interest, no credit checks, just simple, flexible repayment. Get up to$250 with Clio and keep grocery shopping stress-free. Depending on eligibility, you can get up to a$250 cash advance. Terms and conditions apply. Granger knows when you're a procurement manager for an office park, you're not managing one building.

17:46You're managing all of them. And to stay ahead, you need to see through walls and around corners. Lights about to fail, filters ready to clog, HVAC on its last leg. If you wait until something breaks, you're already behind. Count on Grainger for quality products, easy reordering, and 24-7 support. Call 1-800-GRAINGER, click grainger.com, or just stop by. Grainger, for the ones who get it done. Hannah, I just Venmo'd you for dinner. Obsessed. I'm literally spending it right now on the lip gloss that's been sitting in my cart. What do you mean spending it right now? You instantly spend your balance with the Venmo debit card.

18:22Stop. Say more. More exactly. The more you do with Venmo, the more you get, like earning up to 5 % cash back with Venmo Stash. Get the Venmo debit card or check out online. Venmo Stash bundle terms and exclusions apply. Max$100 per month. See terms at venmo.me slash stash terms. Venmo checkout not available at all merchants. The Venmo MasterCard is issued by the Bancorp Bank N.A.

18:49There's more we need to know today. Let's get to the briefs. Intel has its own Computex Week announcements. The company told ARS Technica that it plans to release its Crescent Island GPU by the end of the year, which will use cheaper memory and cooling than NVIDIA and AMD chips. Crescent Island, built on Intel's XE3P architecture, is air-cooled instead of liquid-cooled and uses LPDDR5, not high-bandwidth memory. This would make it cheaper and more power-efficient. It's designed to speed up the inference part of AI, where a system answers your queries. Intel also announced the Xeon 7 Diamond Rapids CPU built on its 18AP node with PCIe 6.0 and 50 % more cores and twice the bandwidth of the Xeon 6.

19:31Diamond Rapids is coming in 2027. And finally, Intel launched the Xeon 6 Plus Clearwater Force data center CPU with 30 % better performance per thread and 30 % better energy efficiency than the AMD EPYC 9965. it's the first chip made on the 18a process for the data set yeah so uh if i'm looking through these announcements uh what intel wanted to make sure people took away is the 18a process is real we're making chips on it chips you can buy uh we are going to compete with amd as the bargain alternative to nvidia in the data center uh you know i mean these are cpus not gpus but still They're comparing it to the AMD EPYC.

20:11They would like to be the CPU choice in the data center, although I think NVIDIA is also the CPU choice for most people in the data center. And they're competing on efficiency. I thought that was interesting as well. When they talk both about the Diamond Rapids and the XE3P Crescent Island, they're talking about we are going to be the cheaper option to run, which is a good position for Intel to be in, even if it's not a traditional position for Intel to be in. We're used to, you know, over the course of our lifetime, Rob, looking at Intel as the most powerful. Yeah, cheaper to run is a good thing right now because in the real world, how do we see this?

20:50We are already seeing the LLMs lower the amount of stuff that we can do on them unless you pay for a account or pay for a higher end account than maybe what you already had. So if there is a company out there that is building processors that these organizations can use to do inferencing at less cost, that's probably better for the end users of the technology. You know what? It's also better for us as consumers who want to buy RAM at a reasonable price, because the faster you can make those data centers efficient, the slower they need to be built, the less RAM that needs to be bought up and the less high bandwidth memory.

21:27if Intel is saying you don't even need high bandwidth memory to make the best use, then you don't have to take up all the capacity to make HBM for people. And that means that you can leave a little capacity for us to make some RAM that we can afford. Tom, you and I are both of an age that when we came up, it wasn't a thing to take RAM out of an old computer and put it into the new computer that you were building. But that kind of went away towards the end of the 90s into the 2020s. Now we're back there. It is not uncommon to see to see people who are like, let me go grab this RAM from this PC from five or six years ago and stick it in this new box just because not because I'm not unwilling to buy the RAM.

22:06I am unable to buy the RAM. It just doesn't exist for me to buy. Yeah. So at any price. So, yeah, so seeing people doing that kind of tinkering is actually, if I want to pull this RAM out, maybe I can pull this card out and put this card here. You're seeing it seems like tinkering with PCs, you know, tinkering with just computers and building your own setups. It's kind of almost in a resurgence area right now because of necessity, because we can't buy the stuff that we want. So let's just go play around the stuff that we already had and MacGyver some stuff into a new PC build. Yeah. Hold that thought, too, because we have some really interesting announcements we're going to get to in a minute that play right into that.

22:41Real quickly, though, I want to tell people about Asus celebrating the 20th anniversary of its Republic of Gamers brand with a special edition gaming handheld called the ROG Xbox Ally X20. Now, granted, you probably thought it was going to be some 20-year-old piece of machinery that had been updated, and this is the same hardware as last year's Xbox Ally X, but it has some improvements. The display now offers a 7.4-inch 1080p 120Hz refresh rate OLED panel, 1400 nits of peak brightness. They also improved TMR joysticks, transformed the D-pad with up to eight-way control, and improved face buttons that now sit flush on the chassis.

23:20And it's bundled with a pair of ROG X-Real R1 Edition 20 gaming glasses. These are augmented reality glasses that work as your display. So they're able to keep up with the frame rates and do decent gaming while you just wear them on your face and you can project a huge, you know, 100 plus inch display in front of you. No price or release date yet on this new anniversary edition of the Ally X20. Asus also announced its first 24.5 inch OLED eSports monitor, has a 540 Hertz refresh rate, announced the ROG Astral GeForce RTX 5090 Edition 20 desktop. This is also a 20th anniversary edition, has a curved AMOLED display on the case to deliver, and I quote, dynamic 3D visuals, custom animations, and real-time hardware data.

24:13So these are still fairly niche devices. This is not something that everybody is going to go out and get, but the people who love them, love them. And just the little upgrades they're doing, this is basically what folks who already had these things are saying. Can you please fix this in the next one? So it's not necessarily a net new device. It's just taking the device that you already had and the little tweaks that a lot of folks have been asking for seems like they've gotten around to doing that. It struck me as a little weird when I first read this, that it was a 20th anniversary edition that's just an upgrade to the previous edition.

24:48But, you know, whatever you need to do to celebrate, right, doesn't doesn't have to always be nostalgia. So, folks, there are lots of Apple rumors kicking around us as we get closer to WWDC. Bloomberg's Mark Gurman says that the Apple smart glasses, codenamed N50, will launch at the end of 2027 with four different styles and working with the revamped Siri expected to come with the operating systems later this year. Gurman says that Apple will focus on competing with traditional eyewear makers as it did with the Apple Watch. Gurman also says we'll get to the new Apple TV and Apple Home Mini devices this autumn.

25:24So the glasses thing is really interesting to me that they're going the route with the watch. Because when you think about the Apple Watch, Apple Watch wasn't competing with, you know, people who already almost all wear watches. They were competing with people who don't wear watches. And can we get them into something that they will right now? I think it's 67, 68 percent of people wear prescription eyeglasses and 92, 93 percent of people wear some type of eyewear. So for Apple to say, no, we're not going to work with any of them, we're going to go build our own stuff. That is they're saying it's government saying it's similar to the to the Apple Watch.

25:59I don't necessarily know that it is because I was a person who would wear a really nice timepiece. But what Apple was looking at with the Apple Watch was that people weren't wearing watches anymore. Let's see if we can put something on the wrist that is kind of net new. Glasses, people wear. So if they look different, if they don't look as cool as what you can get from Zenni Optical or, you know, any, you know, the Ray-Bans or anything like that, people will notice that. So I just I wonder how that's going to work for Apple. Well, don't forget, too, that they partnered with Hermes right out of the gate to try to grab a little bit of that market for the watch as well.

26:31Because I think the other side of this is Apple is not going to try to compete with Meta. Uh, what, what German I think was trying to say is when the Apple watch came out, they didn't try to compete with Pebble. They didn't try to compete with the sort of nascent smartwatch market that was beginning to grow. Even fossil and places were starting to do things or Garmin. Uh, they, they said, look, we're going to make a nice looking smartwatch that anybody would want. Even if you've never thought about a smartwatch before, uh, to your point, even going after people that aren't wearing watches anymore.

27:03And I think that's what they're doing with the glasses. Although Meta is very stylish. They've gone with Ray-Bans. They've gone with Oakley. And you see Google going with Gentle Monster, going with Warby Parker. Apple's banking that they can design something just as good as those. I'm curious if they do a partnership like they did with Hermes, at least for some models of Apple glasses. But it's it's either Apple hubris or typical Apple being right to say, like, yeah, we're good at design. You're going to you're going to love these. You're going to think they're they're good looking and you're going to want to wear Apple as an Apple brand the way you do an Apple watch.

27:44That will be the interesting thing to see. I can't call it. I don't know because Apple is generally pretty good at doing these things. But Meta is not going to stop partnering with Ray Ban and Oakley and whoever else, you know, they come up with. Google is not going to stop. So the fact that these don't come out for a year or more from now and you still have other folks that are coming out with, you know, with some later this year, definitely some in early 2027 just makes me wonder, Apple, how are you going to navigate this? Because if you're doing something that is markedly different from everyone else, maybe it won't be.

28:17Maybe it will be, you know, it's just it's just interesting to see. Yeah, it implies this is not a prediction, but it implies that Apple has figured out something that will be noticeably better than the current glasses. The current glasses are usually like, wow, they look just like regular glasses, but they do still have visible lenses that you can see in the corners. If you if you look, Apple may try to say you can look all you want. You're not going to be able to see our lens like that. That would be the kind of thing that they would have to have to say, this is why we took so long to bring them to you because these just look like glasses.

28:55You can't tell the difference. I just so happen to be in a fairly enclosed place with lots of people wearing Ray-Ban mattress. You don't even notice after like the second and third pair. You just don't. You notice when the light's on. But when they're just being worn as glasses, you don't even notice anything. I mean, you really have to look to see. It's sunny outside. It's bright. If you don't have glasses on, you can't really see what other people's glasses look like. But I did notice that. There were hundreds of people that I have seen recently wearing Ray-Ban Metas, a few wearing the Oakleys, and you just don't even notice anymore.

29:29It just looks like sunglasses. I think they are a little more mass consumer already than the smartwatches were at the same stage. Right. I'm a little curious in how much this metaphor holds up. Well, to give consumers a more affordable CPU option, we were just talking about this earlier, AMD is re-releasing the four-year-old Ryzen 7 5800X3D for the AM4 motherboards. Goes on sale June 25th, starting at$349. That's$100 less than when it first went on sale in 2022. AMD says that this is a celebration of an anniversary, the 10th anniversary of the AM4 socket. So they're releasing a four-year-old CPU to go with it.

30:13AM4 socket launched, of course, back in 2016. AMD also planning on launching the Ryzen 7 7700X3D desktop CPU. That's for budget gamers. That'll have a suggested retail price of$329. So a few dollars less,$20 less than the other one. That should go back on sale July 16th. And this one's not a re-release. AMD is releasing the Radeon RX 9070 GRE edition globally. This one launched in China last year. It is a little bit less than the other 9070s. 49 RDNA for compute units instead of the 56 or 64 you get with the other 9070 models. Also has 128 gigabytes of video memory instead of 16. It's available now starting at$549.

31:04So just a bit of a step down. So here's the question. What is the comparison in price? It's a step down, but is the price also a step down? Yes, the price is also a step down. Exactly. So, so yeah. So, you know, you know, giving price points to different folks, because like I said, when you're building these PCs, when you're building these computers, they can get expensive very, very quickly. So just a hundred,$150 difference can make a world of difference to somebody who didn't have the extra$150 to spend to go to the higher end. Yeah. And by the way, I realized I said 128 gigabytes of video memory, which would be crazy.

31:4112, 12 gigabytes of video memory instead of 16. But yeah, these are cheaper. I mean, actually, I shouldn't say that.

31:53The GRE costs$549, which is actually list price the same as the Radeon 9070 without a GRE or anything else after it. But you can't find that one for$549. And the expectation is that they're going to raise the list price of the 9070 to something around$700. and then$549 becomes the price of the GRE edition and that you will be able to get it for$549. So thanks for following up on that because I wanted to make sure that I got that right. Well, folks, we do live streams here at Daily Tech News Show and you can catch them by becoming a subscriber over at youtube.com forward slash Daily Tech News Show.

32:42This episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi-agentic AI. They already deployed one. It's called Chat Concierge, and it's simplifying car shopping. Using self-reflection and layered reasoning with live API checks, it doesn't just help buyers find a car they love. It helps schedule a test drive, get pre-approved for financing, and estimate trade-in value. Advanced, intuitive, and deployed. That's how they stack. That's technology at Capital One. When you're a maintenance engineer in a beverage manufacturing plant, you keep production lines moving and quality on track because there's no room for slowdowns.

33:21With Grainger's vast selection of high-quality motors, sensors, belts, and hard-to-find parts, you can get what you need fast and all in one place so nothing gets in the way of getting the job done. Call 1-800-GRAINGER, click grainger.com, or just stop by. Grainger. For the ones who get it done.

33:44okay we got some other headlines that are going to make you look smart pay attention anthropic plans to give the eu security organization anisa access to mythos becoming the first european agency to get access the information's sources say meta is working on releasing four more models of smart glasses to our point by the end of this year and also developing a smart pendant that it will begin testing over the course of the year alienware announced what it claims is the world's first 39 inch 5k OLED monitor featuring LG's RGB stripe tech i mean makes text and small things look better Dell is bringing back the XPS 13 as a MacBook Neo competitor so many companies trying to take on the MacBook Neo now the XPS 13 has an aluminum chassis weighs 2.2 pounds is 12.7 millimeters thick but starts at$100 more than the MacBook Neo.

34:40I'm excited. $699. I'm super excited to see this one. I will say as excited as I was to see the MacBook Neo, which I love, by the way. Yeah, yeah. I'm very excited to see this one come out. We'll have to do a side-by-side comparison, I think. Absolutely. Strava is restricting access to its website and other data, only allowing authenticated users to access it in an effort to combat LLM scraping. And I would imagine other companies are going to do the same. Yeah, yeah. Twitch announced that next month, creators will be able to stream in both vertical and landscape formats simultaneously, something you can do on YouTube already.

35:15DuckDuckGo has added an option to make its No AI version of Search Engine the default on both Chrome and Firefox with extensions. Ah, through extensions. But yeah, I guess that No AI version of DuckDuckGo has gotten so popular it was worth throwing an extension out there. Yeah. Well, folks, we end every episode of DTNS with some shared perspectives. And today, Allison has a question on the use of the term underemployed when we were discussing the job losses related to AI last week. Yes, Allison was reading the transcript email that comes out first and then listened to the podcast episode, hoping for some clarification on the AI is taking our jobs numbers.

35:53Allison writes, overall, the audio did clarify quite a bit. Where I'm hung up is on underemployment. In the audio, you explained that this statistic is about people who are doing jobs that don't require the college degree they received. Here's what you said in the transcript. The unemployment rate for recent college graduates rose to 5.6 percent, while the underemployment rate, the share of graduates working in jobs that typically do not require a college degree, reached 42.5 percent, its highest level since the COVID pandemic. that isn't all caused by AI, but it isn't slowing down. And then Allison comes back in and says, the thing I'm trying to figure out is, does this 42.5 % mean that nearly half of all college graduates end up working in a field where they didn't require their degree?

36:41I know it says that this was its highest level since the COVID pandemic, but what is it normally? Assuming this includes people who get a degree in poetry and end up working at the mini mart in a gas station, and normally the stat is 38%, I'd have a better grasp of what the stat change signifies. And so I wrote this back to Allison, but yes, that 42.5 % means that nearly half of all college graduates end up working in a field where they didn't require a degree. So four out of 10 recent college graduates, as defined by the New York Fed, as those aged 22 to 27, are working in a job that typically doesn't require any kind of bachelor's degree.

37:22So this isn't just about like I majored in poetry, but I'm working in a job that doesn't require poetry. It's a job that doesn't require a college degree at all. According to the New York Fed, over the past three decades, the unemployment rate for recent grads has consistently hovered between 38 and 44 percent. And it peaked near 47 % in the aftermath of the Great Recession around 2014 due to a lagging job market for youth and then spiked again during the COVID pandemic lockdowns as well. So a 44 % spike was above 42 % in 2020. And we're coming back up to those same levels. So you can think of it that way is it's harder for a graduate to get a job.

38:05I think it's less instructive to focus on that number itself and just note that it is hard to get a job that requires your degree as it is as hard to get a job that requires your degree as it was in 2020 and in 2014. Yes. And one of the other things that I was thinking about reading through this is that the first graduating class, you just graduated in 2026. They are the first, the last class to graduate that did not know about AI before AI became what is being. Everyone who's come after 2026, they at least had an inkling that it was going to be something. So that's part of this as well. I think that one of the things that maybe a lot of people don't realize is that most jobs don't require a college degree.

38:49However, college degrees are immensely helpful in getting some of those jobs that don't actually require them. They will actually use those degrees to say, oh, this person actually finished a four year degree program. They don't have any experience, but I might give that person a little bit more weight over the person who's just coming out of high school with no experience in this field as well. So don't want that to sound like you don't need a degree. You know, in a lot of cases, you don't. And I'm not one of those folks that says you just go to college no matter what just to get the degree. But I do think that sometimes degrees get a bad rap.

39:19They do help you in getting jobs, even when those jobs do not require a college degree in many, many, many cases. And definitely go listen to last Tuesday's episode if you missed it, because we were teasing out what the effect of AI on jobs is. And it is not nearly as widespread as people seem to think, at least so far. And so Jason and I dug into the numbers and said, well, we think we can see an effect in this particular area, which is entry-level workers. But there are other reasons that entry-level workers are having a harder time getting a job now than others that are unrelated to AI, that are just economic trends that are happening right now as well.

40:00Yeah, absolutely. And then one of the things I would like to see, and it will take time for this, is that because there are now folks who are entering college understanding that AI is a big thing, does that change what they go into? So it's just going to take time. It's going to take another year or two or three or four to kind of see how those numbers spread out. Are there less people getting the quote unquote technical degrees that four or five, six years ago you thought that if you just went into anything tech, you were golden when you came out of college? I would imagine that the case is going to be that those numbers have probably gone down a bit.

40:29But we'll have to see the data to do. Yeah, absolutely. Things are moving fast. Well, folks, we want to know what you're thinking about. So if you've got some insights into a story, please, please, please share it with us. Feedback at DailyTechNewsShow.com. Big thanks to Allison, who's always good at asking the good questions like that for contributing to today's show. Thank you for being along for Daily Tech News Show. You're the folks who keep us in business. You know that. If you're not a patron already, you can become one. Patreon.com slash DTNS. Talk to you soon.

41:07The DTNS family of podcasts. Helping each other understand. Diamond Club hopes you have enjoyed this program.

41:20This episode is brought to you by Capital One. Capital One's tech team isn't just talking about multi-agentic AI. They already deployed one. It's called Chat Concierge and it's simplifying car shopping. Using self-reflection and layered reasoning with live API checks, it doesn't just help buyers find a car they love. It helps schedule a test drive, get pre-approved for financing, and estimate trade in value. advanced, intuitive, and deployed. That's how they stack. That's technology at Capital One. Security program on spreadsheets, new regulations piling up, and audit dread? It's time for Vanta.

41:56Vanta automates security and compliance, brings evidence into one place, and cuts audit prep by 82%. Less manual work, clearer visibility, faster deals. Zero chaos. Call it compliance, or call it compliance. Get it? Join the 15 ,000 companies using Vanta to prove trust. Go to Vanta.com slash com.