Nvidia's RTX Spark superchip partnership with Microsoft represents a strategic timing opportunity where Nvidia's dominant position in AI hardware allows them to influence Microsoft's Copilot Plus PC requirements, potentially enabling GPU-based AI processing in Windows devices and setting the stage for Windows 12's AI capabilities.
Approfondir
Prérequis
- Pas de données disponibles.
Prochaines étapes
- Pas de données disponibles.
Approfondir
First Ring Daily 1972: The Spark has ArrivedAjouté :
Hear me out. I think Microsoft got really lucky.
Just just hear me out.
>> Like from a timing perspective?
>> From an everything perspective. So >> Okay.
>> For those who aren't aware, you're probably aware but aren't aware.
Anyways, Nvidia announced their RTX Spark superchip as they call it, which is going to power a line of Windows devices.
However, this is why I think they got real lucky.
Paul, you and I both know this initiative did not start 3 months ago.
This was probably 3 to 4 years in the making.
>> Easily, yeah.
>> At that time Nvidia was a much smaller company. AI was still like sort of happening but not not like has been for the past 18 months.
>> Right.
>> So they they start working on this thing. All of a sudden, Nvidia does a hockey stick of growth. AI explodes. And Windows is kind of like floundering a little bit. It's like it needs it needs some mojo. And then here comes the only company, in my opinion, who can actually twist the arm of Microsoft to do whatever it wants.
Shows up and says here's a new superchip for your laptops running this OS. You're welcome.
>> Yeah. Um Yep. I definitely I there is an element of this that is similar to Qualcomm entering the PC market. Which beyond the obvious, right? Like ARM-based chips, etc. etc. Which is why would a company that has this ginormous business over here work on this little sliver of a nothing over here. And I'm I got to be honest, I'm still a little unclear on this. I I know from you know, but for whatever reasons like Macs, especially Mac minis, really popular with this local AI stuff right now.
I don't like I've said, I don't think they're particularly, you know, well adapted for this or whatever, but they are, you you the whatever. It's very popular.
Um Nvidia dominates in the data center, obviously.
Um there there's these things happening this year which are just open questions, right?
And some of those things might be answered some of those questions might be answered as recently as this week because we have uh or as soon as this week because we have build happening. But, for example, Snapdragon seed low-end Qualcomm chip.
Um 8 gigs of RAM.
So, not a Copilot Plus PC unless something changes, right? And then this thing, which by the way, you know, there's not a lot of technical details, but it's pretty clear that the power of it is coming from these CUDA cores, right? GPU basically, integrated on the SOC. Not from an MPU, not that there might not be an MPU if this is in fact a MediaTek uh partnership and that they have powerful Copilot Plus PC class MPUs, but this is Nvidia. They they might want to go with the GPU, in which case you just mentioned something I think is very pertinent, which is they have the power to make Microsoft change their mind and do different things.
>> Yep.
>> And one of those things is what makes a Copilot Plus PC a Copilot Plus PC, which is that thing >> Here's my next point about for years.
>> Here's my next point. You're going to love this. Go find Copilot Plus PC mentioned anywhere in these announcements cuz I can already tell you it's not.
>> It's not as currently um >> Not even in the Surface one, though.
Does not mention on the Surface Laptop Ultra. They >> Yeah.
>> There's two >> I right. No, I know. I I looked.
And that's that's why I said I don't think the MPU is there.
And or >> Here's >> they're going to I think they're going to change it. I think they're going to either get rid of this brand or just expand it to include GPUs.
>> So, Ryan Shrout responded to my tweet about it. And what did he say?
Um I think he said that Microsoft confirmed that it is a Copilot Plus PC and has an NPU. So, it's probably on the MediaTek.
>> Okay.
>> Yeah, I mean, so the the latest uh flagship MediaTek processor not including the whatever this is. I mean, not not including this. I mean, the things that they have made for uh Chrome Chromebook Plus devices are what I would call Copilot Plus PC class chips, right? They have I don't remember what it is, but let's call it a F50 NP, you know, tops NP or something like that. It They're they they do qualify.
So, if this is in fact a MediaTek partnership, which I it is, right?
Uh it does make sense. Of course, this would just be on the die, but that's not enough. Like I again, to your point, Nvidia is powerful enough to get Microsoft to change its mind. And one of the things that has needed to change for 2 years now is allowing those with GPUs to get the Copilot Plus PCs. So, why would you prevent this?
>> Yep.
>> You know?
And for that stuff to run against the GPU. So, one of the things that uh Pavan or one of the things that Microsoft through Pavan Davuluri uh has kind of hinted at on Twitter uh at a build is that there'll be some developer Windows stuff happening, not a new version of Windows, he said. But, I think it's going to be this, you know, if you're targeting local AI, that Microsoft, you know, this. I mean, you've done this kind of stuff a little bit. Like they they have uh models and technology that lets, you know, that does a sort of um auto routing to whatever resources you have if you're using local on-device AI. But, the Copilot Plus PC stuff could very easily I could run on a lot of it could run on a CPU, but uh could very easily run on a GPU, too.
And I think they're going to enable that for these computers because that's what will make it like a supercomputer, right? Like the Um if you do like a something simple like um super resolution on a photo or something, maybe it takes 30 seconds on a MPU and on an Nvidia GPU it takes 10 seconds or something. You know, maybe. I you know, we'll see, but um I think they the thing you said that I keep going back to it is like, yeah, they they're powerful enough that Microsoft will do their bidding and >> this needed to happen previously, but um hopefully it happens broadly now.
>> I think this also, and I know they said they're not going to announce this at Build, but I also think this sets up Windows 12 for probably next year, is my guess, right?
They're going to go to Build and they're going to say, "Here's our new APIs and all this stuff."
>> [clears throat] >> It'll eventually ship. We're working on it. Blah blah blah. It'll take them longer than >> This is the wrong year to announce a Windows 12, especially if it has new hardware requirements at a time when we can't get RAM or storage or whatever it is.
Um because you know, like you know, you can the 16 gigs of RAM thing is a minimum Copilot+ PC makes sense to me. It's fine.
>> Mhm.
>> Um I think you need more than 256 gig of storage or whatever, but >> most modern >> you know, the MPU feels arbitrary, but see, I feel like the uh the the what do you call it? The baseline, I guess, should be tops.
Not MPU tops.
>> Oh, I agree.
>> Copilot+ PC, right? I think and and maybe that's the >> the next way >> when we go to 12.
>> Right. So, I think it'll line Here Here's my guess.
>> Mhm.
>> Next gen Xbox comes out next year.
It's going to have some new gaming, whatever, unified OS thing. They've got this AI stuff baking. They've already got the story starting to write itself.
It's like the best OS for local agents and and they'll have the proper API. I don't know what the hardware requirements will or will not be. Um backwards None of Nobody knows right now.
>> Nobody knows, yeah.
>> But I think next year it starts to make sense that that is when Cuz let's let's be real here. This stuff will launch in the fall. It'll take time for it to get any sort of market penetration. So you give it a year and then by this time next year, let's just say the fall of next year, these devices, if things go well, could be fairly standardized, if you will. Like Meaning like walk into Best Buy and there's one for $500 all the way up to $5,000.
>> I think the >> There you go.
>> the Nvidia Look, this these computers, like and the Surface Laptop Ultra whatever it's called, are going to be incredibly expensive.
They're going to be very niche-y kind of devices, but they're going to appeal to people, developers especially, who are currently getting Mac Minis or what are trying to.
Um and want to target the same kind of CUDA APIs that Nvidia offers up in the cloud, right? But locally. And I think that's going to be either their appeal, but I but to your point uh I keep saying for some reason. Um I think if you assuming Windows 12 is the next thing and that's what goes down, I you know, they make that more of a mainstream kind of a thing for that, if that makes sense. Um and are gambling that the component crisis will ease or pass, you know, by the time that thing would probably launch in late 2027, right?
>> Here's my other guess, too.
Related to Nvidia bending the arm of Microsoft. Nvidia wants does not want AI PCs known as Copilot Plus PCs.
Despite the fact that it's terrible branding, they want to establish that RTX Spark means AI PC.
>> AI.
Right.
>> And that's why we won't see Copilot Plus PC branding for any of these, like Lenovo or whatever. Nvidia's like, "If you're using RTX Spark, you're calling it an RTX Spark laptop, not a Copilot Plus PC.
>> Like what cuz we need more confusion in this marketplace. Like um a lot of HP does this. I know I'm I think Lenovo does as well. Like a lot of companies still just use the term AI PC.
And you can kind of you can semantic it down if you want to. You can be like, well, you know, Intel in particular sells whatever line of GP or CPUs rather that don't have a powerful enough MPU to meet the spec for Copilot Plus PC, but it's still an MPU and you can still do things like Windows Studio effects to some degree or window what's it called? Yeah, Windows Studio effects to some degree.
It's an AI PC and you're like, yeah, okay, you know. And I guess you know, in you can make a really good case and I think this is the Nvidia case that um uh desktop computer in particular, but modern laptops with good dedicated GPUs, Nvidia GPUs, um are AI PCs because those things have don't have dozens of tops. They have hundreds of tops, you know, performance for for AI.
>> I really hope that I didn't put it together initially, but I hope that you're correct in that Nvidia's like, look, these anything with like a 5000 series GPU has more than 40 tops. And already does.
>> A lot more in most cases, yeah.
>> And you're you're artificially limiting us. If you want us to go to market with this, you're going to remove that barrier. And maybe that's a Windows 12 thing, whatever. But as long as as long as like that's part of the why reason why I think a Windows 12 needs to exist is because the branding just needs to be cleaned up because >> Yeah, you all have to say And and also just like I said, I these things are going to be so expensive, right? So they there's a bar that's being cleared here that Microsoft's trying to establish for a Copilot Plus PC anyway, right? These are premium PCs. These are going to be ultra premium PCs. These are going to be This is going to be an interesting gut check moment for people cuz I think people trust Nvidia enough that they may get over their little, you know, Windows on ARM thing, but um you know, these are like Windows in our like portable workstations basically. So, this is an interesting new area, but you can't really make that complaint about you know, price positioning, whatever, but Windows 12 is an assuming we're you know, we're just speculating, but I mean, is an interesting opportunity to bring those capabilities mainstream. In other words, you have a computer desktop, laptop, whatever dedicated GPU hundreds of tops and now you get those capabilities just by upgrading to Windows 12.
>> Yep. I agree.
>> Cuz you can't really upgrade the Copilot Plus PC, you know, for some reason, but >> They'll probably come out, if I had to guess, some sort of baseline requirement that has eight gangs eight gangs eight gigs of dedicated video memory.
>> Yep.
>> And then anything if you have that, then you will get the you know, the AI features.
>> I mean, I the one thing like I the one thing I kind of feel bad about here is the benefits of an NPU are actually real. I mean, you know, some people think this is made up or whatever.
That's not really the issue. Like, if you're on a laptop, especially like an ultra-thin laptop on battery you know, you want the most efficient thing. I mean, no doubt about it.
But if you spent money on like a gaming laptop or portable workstation, whatever it is you have some hawking GPU in there not being able to access these advanced features that would easily run easily.
And you're like, well, yeah, but you're going to consume more electricity. I don't care. If I cared about that, I would never bought this thing in the first place.
You know?
>> Yep. Big day for Windows.
Big day for Windows.
>> Yeah.
Yeah.
Vidéos Similaires
BREAKING: Microsoft’s New Image Generating Model Beat Out GPT 1.5 and Nano Banana 2
aimmediahouse
122 views•2026-06-03
Are AI deceiving us? | Roman Yampolsky, Gleb Solomin #AI #science
shortsGlebSolomin
1K views•2026-06-02
Nvidia Bets Big On AI PCs | New Chip To Power Windows Laptops | Technology | AI Updates | N18S
cnnnews18
3K views•2026-06-01
AI Doesn't Create Bias — It Inherits It
UXEvolved
176 views•2026-06-01
Distributed Inference Challenges Explained #shorts
alexa_griffith
466 views•2026-05-31
[한글자막] OpenAI @ Replay 2026 | OpenAI는 Codex로 개발 방식을 어떻게 바꾸고 있을까요?
TechBridge-KR
1K views•2026-06-03
Starting & Test Driving JAKE'S Abandoned BUS from Subway Surfers | POV Restarting
RestartGaragePOV
4K views•2026-06-04
Building the Future of Voice-First Sovereign AI: Sarvam & NVIDIA
NVIDIA
3K views•2026-06-01











