NVIDIA teases “new era of PC” ahead of N1 and N1X laptop chip announcement
Posted by PaiDuck@reddit | hardware | View on Reddit | 300 comments
Posted by PaiDuck@reddit | hardware | View on Reddit | 300 comments
Qwen_os_has_died@reddit
Mfer can't even ship one more 4090 to me.
VenditatioDelendaEst@reddit
Well yeah, it's been superseded, no?
doscomputer@reddit
in terms of market segment and performance, no. The 5080 and 5090 do not act as a direct replacement in either regard
VenditatioDelendaEst@reddit
The 5080 is "enough GPU for any gaming application" market segment.
The 5090 is "the fastest GPU money can buy" market segment. This is where the 4090 was, and no longer is.
fullmetaljackass@reddit
No, it's what you settle for when you can't afford an RTX Pro 6000 Blackwell segment.
CrowdGoesWildWoooo@reddit
Gaming GPU has better price to performance compared to the Pro line significantly. Only the xx90 lines has enough VRAM for AI.
Lefties_TheWorst7331@reddit
They're same performance speed as the 5090 tho aren't they? And 3x vRAM that is ECC.
VenditatioDelendaEst@reddit
The benchmark charts do not have the RTX Pro 6000 Blackwell at the top of them, and that makes all the difference in the world.
fullmetaljackass@reddit
TPU? Seriously?
But yeah, it's not on the top of the list because they didn't test it, so I don't understand what point you think you're making.
If you refer to benchmarks that actually include the RTX 6000, you'll see that it has a clear lead in every test, with the exception of CP2077 @ 1080p w/ RTX off, where it appears that the game is CPU bound, and the 5090 takes the lead by 0.3 FPS.
VenditatioDelendaEst@reddit
TPU is the best English-language hardware reviewer, but here it it doesn't matter which one you pick.
Techspot/HWUB: No RTX pro 6000
PCGH: No RTX Pro 6000
GamersNexus outside of the one time they benched a workstation card as a bit: No RTX Pro 6000.
I believe you.
Please recall that I did not say the 5090 is the fastest GPU money can buy. I said it covers that market segment.
That means that approximately everybody who dreams of building a gaming computer with the fastest GPU money can buy is thinking of the 5090. They are not thinking of the RTX Pro 6000 Blackwell.
-Sliced-@reddit
RTX Pro 6000 Blackwell isn't really faster than 5090. Its main advantage is the 96GB RAM - which then does translate to faster performance for local AI if the model is larger than 32GB.
nmkd@reddit
Let me see you play Alan Wake 2 with path tracing at 4K on a 5080
Ecstatic-Art5745@reddit
my dual 4k 57" is cooking my 5080 even on low graphics in some lesser optimized games like squad ...so no 5080 is not "enough gpu for any gaming application" pleb.
DreiImWeggla@reddit
My 4K Monitor and 5080 beg to differ on the first claim
SETHW@reddit
Also VR, any high resolution high framerate application
imaginary_num6er@reddit
Yeah. That's the 5070 having "4090 performance"
faisalkl@reddit
I wouldn't be surprised if there was some dumb attempt to undermine Linux.
Idkwatnam@reddit
As the other reply mentioned, the GB10 already runs Linux but I don't think you'd be entirely wrong. I have a feeling ACPI support will there on laptops as it is on the GB10 boxes BUT that device trees will need to be used to get full hardware functionality (which will bring us to the same situation as Qualcomm's Snapdragon laptops running Linux):
The GB10 seems to lack a lot of features in its firmware (or OS?) that a modern laptop uses. It lacks soundwire support (what modern laptops use for speakers), ISP support, and even eDP support at this stage. Basically anything that a regular desktop computer needs, the GB10 just doesn't have enabled in firmware yet seemingly but is on the hardware, or maybe it just doesn't work with ACPI and would be enabled by using DTB (since there's probably already drivers for these things from the Jeston boards but those use DTB instead of ACPI). See section 4.2.7. Performance Specifications (page 19) in this PDF : https://docs.nvidia.com/dgx/dgx-spark-porting-guide/dgx-spark-porting-guide.pdf
DerpSenpai@reddit
That makes sense. QC engineers said they couldnt go full ACPI yet because of the custom data they need. Passing custom data on the ACPI on Linux means that a kernel that doesn't know how to read that will not work properly anyway
Idkwatnam@reddit
Yeah, ACPI seems like it's not feasible to support anything other than the basic power and peripherals unless your company/chip is already invested into ACPI like AMD/Intel. Why write new ACPI detection drivers for special hardware where there are already drivers that work off DTB.
I think a clear example of this is the Radxa Orion O6, that board is ARM SystemReady certified which means you can boot with ACPI, but under ACPI you don't get access to all of the hardware. Iirc sound doesn't work? and the GPU doesn't work because Mesa expects Mali GPUs to be on DTB devices apparently.
Hot-Software-9396@reddit
mfers will bring up Linux any chance they get
Malygos_Spellweaver@reddit
How? Nvidia uses their own distro. Also https://old.reddit.com/r/linux/comments/1tknraq/phoronix_just_posted_a_pic_with_jensen_huang/
From-UoM@reddit
N1X is the GB10 chip found on DGX Spark. And the DGX Spark only runs Linux.
So expect the N1X to have very good Linux support out of the box.
Artoriuz@reddit
The most exciting thing about this to me is the great Linux support.
12Danny123@reddit
Not sure since ARM doesn't use ACPI which is standardized across all X86 device but instead uses Device trees for each device, this means you need to create drivers for each device for Linux to run and this is also assuming the bootloader isn't locked down.
mhall119@reddit
Arm has supported ACPI for a while, it's just been more widely used on servers than devices (which still prefer Device Tree)
VenditatioDelendaEst@reddit
If this is real it has real potential for running x86 games with FEX.
Idkwatnam@reddit
I mean it's just a GB10, so you could also just look at people's results from running games there (which there are surprisingly little reviews of oddly, guess everyone got paid to ONLY talk about AI performance or something)
Nimbus420i@reddit
Lmao, Nvidia and good linux support? Sorry but that’s far from reality, I will say that it has gotten better lately. But it’s no where near good.
Artoriuz@reddit
Nvidia supports Linux orders of magnitude better than Qualcomm and Apple, which are the points of comparison here.
TitleEfficient3207@reddit
before I read the article - its new APU...
your_mind_aches@reddit
Yeah but it seems like this is an attempt to try to make a better ARM transition. Honestly, I hope it goes well.
Just like I'll root for Intel with GPUs, I'll root for Nvidia (and Mediatek lol) with CPUs.
TitleEfficient3207@reddit
Its a new APU.
To make sure AI OS can get off the ground.
your_mind_aches@reddit
Yes, we know this know. They're trying to get creators interested with those demos because it's not actually generative AI output, it's the same output as any other application, but an agent using the actual tools like a human does.
We can hope the tech ends up reaching lower end systems eventually, like I said. Right now it's basically creators only.
ResponsibleJudge3172@reddit
Yes, makes sense to me. If you like completion. Get more players in GPUs AND CPUs. Not just one
pmjm@reddit
Would be nice to get one of these in a new Nvidia Shield type device.
rohit275@reddit
I would kill for a Shield TV that supports 4K 120Hz honestly. I have the original 2015 and it is still a good device, just missing the AI upscaling and dolby vision features of the 2019 version but otherwise almost the same and hard to justify an upgrade.
GazelleInitial2050@reddit
Yeah definitely not worth the upgrade.
I love my 2019 shield, but it's showing it's age.
Things I want in a new one:
EnglishBrekkie_1604@reddit
It’s crazy because Nvidia has a chip that can DO ALL OF THIS, the switch 2 chip! It’s just stubbornness stopping them from doing what they did last time with the switch 1 and original shield.
Moscato359@reddit
Closest you get is apple tv right now
rohit275@reddit
Almost, but it can't do 4k120hz either. Also the whole Apple walled garden thing.
It is a really nice streaming box though, and no ads all over the interface is genuinely very nice.
Moscato359@reddit
4k 120fps video is almost non existent though
rohit275@reddit
You're not wrong ... but would be nice for gaming purposes if you have a gpu good enough to push those frames.
I do concede it is a very niche use case for 99% of folks out there haha
rohit275@reddit
If such a device existed I would instantly buy tbh
Earthborn92@reddit
Shield is still so good after this long.
A Shield 2 would be an instant purchase for me if it fixed all the deficiencies like the lack of AV1 decode
joulesFect@reddit
The shield is an amazing device, but I'm so sour they removed gamestream from it haha
rohit275@reddit
I was shocked when I realized they did that. It was like the main reason I even wanted to get the thing back in 2015 (back when they cared about such things) over other streaming boxes.
But sunshine and moonlight is a genuinely nice replacement using the gamestream protocol so it's not the end of the world.
noiserr@reddit
This thing is like $2000+. Might as well get the Seam Machine it's way cheaper.
Sea-Sir2754@reddit
Maybe in a few years when they have a few hundred thousand of these old chips to use up. Not when it's the groundbreaking new chip going into laptops.
RootExploit@reddit
I am not familiar with either chip, what's the excitement about?
pmjm@reddit
They're ARM cpu's with Blackwell graphics.
alabasterskim@reddit
Do you mean a Shield TV? Or Tablet?
pt-guzzardo@reddit
Everyone forgets about the OG Shield. :(
sowee@reddit
I'm glad I did, it really was an ugly specimen!
BarKnight@reddit
Yes
alabasterskim@reddit
Haha, I only asked because I feel like there's gotta be some secret agreement between Nvidia and Nintendo for Nvidia to not compete with them on gaming handhelds. The Shield Tab got discontinued like a year after the Switch and we never saw another relaunch while they clearly still had the chip know-how to make it happen.
avengers93@reddit
For real
OttawaDog@reddit
The era where you pay $3000 for a laptop that has to emulate x86 for most software?
CalmSpinach2140@reddit
It won’t be $3000 for the starting price, more like $1500-$2000
OttawaDog@reddit
Maybe for the Non-X chip, but that has a tiny useless GPU less powerful than an RTX 5050. So why would anyone want an emulated CPU and weak GPU, for still more money that you get a real x86 CPU and more powerful discrete GPU??
Past_Bathroom5568@reddit
I mean macOS did it too 🤷♂️
OttawaDog@reddit
Apple commits though. MS dabbles. MS first launched an Arm version of Windows way back in 2012 with Windows 8... It's still a second class citizen today.
Apple only first launched an ARM Mac in 2020, and Mac x86 is like a distant memory now.
Past_Bathroom5568@reddit
Windows has the decency to keep supporting x86 and arm. Meanwhile macOS essentially turned all Intel Macs into a paperweight overnight. Won’t be buying another Mac because im sure they will release some new bs and brick more perfectly good hardware
Ecstatic-Art5745@reddit
lol wut
they continued to support for many years and the performance is night and day better on arm with the m chips vs the old intel trash
I just retired my old intel powered 2017 macbook late 2025....it worked fine till then. There was no need for it to live longer.
dagmx@reddit
Explain how they’re bricked and paperweights. Do you think they actually just stopped working?
WhiteNamesInChat@reddit
You mean over 7-10 years? You can be mad about it, but that's far from "overnight".
Old-Board1553@reddit
Ok but macOS moved completely to ARM.
WhiteNamesInChat@reddit
Over the course of seven years, sure.
Old-Board1553@reddit
Bro it didn't take that much 😄) They only announce they cut support for macOS and Intel chips.
_FlyingWhales@reddit
x86 is absolutely ancient. Good riddance.
WhiteNamesInChat@reddit
ISA doesn't really matter for general purpose computing. Everything gets decoded to micro-ops anyway, and 90% of what you do is load, store, arithmetic, compare, and branch.
_FlyingWhales@reddit
x86 has decades of legacy instructions causing the decode step to be inefficient. What is even more important however is the huge challenge of securing and validating an instruction set of this size. They tried fixing it with x86S but abandoned it last year.
WhiteNamesInChat@reddit
Decode and number of instructions really don't move the needle a whole lot.
If you're building a new platform from scratch, then of course you don't start with x64, but it's not like we'd be in a much different state if Intel and AMD had been implementing ARM architectures all these years.
Gaycel68@reddit
Fuck most software
OttawaDog@reddit
You in a hurry to make NVidia richer?
Gaycel68@reddit
I really want x86_64 to die
ViniCaian@reddit
Never going to happen. Especially now that the worst of Intel is behind them and AMD is ever more competitive.
Gaycel68@reddit
Why would you think Intel is happy with x86? They tried to kill it at least once before.
ViniCaian@reddit
Decades ago, yes. Didn't work out too well, and from the amount of effort they put into AVX10 and the stablishment of the x86 consortium, seems like that's not something they want to dwelve into again.
bobj33@reddit
Intel tried to kill x86 in 1981
https://en.wikipedia.org/wiki/Intel_iAPX_432
Intel tried to kill x86 in 1989
https://en.wikipedia.org/wiki/Intel_i860
Intel tried to kill x86 in 2001
https://en.wikipedia.org/wiki/Itanium
I may have missed a few other chips.
ViniCaian@reddit
iAPX was super cool as a product of its era, sometimes I wish it had taken off in some capacity
The fact that it died despite the success of OOP as a programming paradigm is unfortunate considering how much of its design was purpose built for that.
Gaycel68@reddit
It’s going to happen. The vast majority of consumer devices on the planet are non-x86 already. We just need more decent Windows ARM laptops, and x86 will go away in its own.
Luggage-Lock@reddit
When did they try to do that? I remember awhile back rumors of ending 16bit and 32 bit support and moving to x86S but was quickly abandoned.
bobj33@reddit
Intel tried to kill x86 in 1981
https://en.wikipedia.org/wiki/Intel_iAPX_432
Intel tried to kill x86 in 1989
https://en.wikipedia.org/wiki/Intel_i860
Intel tried to kill x86 in 2001
https://en.wikipedia.org/wiki/Itanium
I may have missed a few other chips.
Gaycel68@reddit
Itanium n 2001
Luggage-Lock@reddit
Oooh interesting, I gotta do some reading up on that one. Thank you!
the_dude_that_faps@reddit
Because they couldn't think of doing what AMD did with amd64.
DrKersh@reddit
it will happen and soon, I don't give x86 more than 10 15 years before everything is arm.
it is more efficient per watt and x86 have a lot of tech debt.
hackenclaw@reddit
and 16GB RAM marketed as Premium high end... with 8GB being the mid-upper end segment.
DerpSenpai@reddit
Most software is ARM native by now except some Pro software and gaming, and in gaming what matters is the GPU in a laptop
TechTechTerrible@reddit
Modern games use avx2 instructions and emulating those instructions is ~30% slower. Depending on the game that might result in a decent hit to performance even with a nice gpu. I'm hoping NVIDIA has enough pull to get some devs to recompile their games for arm.
DerpSenpai@reddit
30% of the X2 Elite is still better than Panther Lake.
The CPU doesn't matter in laptop gaming, 30% of the X925 in the N1X is laptop Zen 4
Verite_Rendition@reddit
I'm not sure why this is getting downvoted. The most commonly used Windows applications are available as Arm binaries. MS and Qualcomm have been putting a lot of work into this over the last several years, particularly on productivity and common consumer software.
https://worksonwoa.com/
The only software that is really an issue right now are performance critical applications without a native binary - which outside of games is few due to the above efforts - and then applications that need to load hardware drivers. That's primarily games again.
DrKersh@reddit
99.9% of the software is not arm native
siazdghw@reddit
This is a straight up lie, unless you're weirdly including mobile apps and counting MacOS apps too.
Windows and Linux have MILLIONS of applications for them over their decades of existence. While it's impossible to give you an exact number, around 99.9% are built exclusively for x86-x64. Software compiled natively to run on ARM is 'new' as there simply was no point for developers to do that previously, the last attempt was the old Surface (RT) and successors which sold horribly compared to the Surface Pro with the Intel chips running full Windows.
Now of course quite a few popular applications have ARM ports these days, but to say "most software" is absolutely incorrect.
DerpSenpai@reddit
Most software people use.
OttawaDog@reddit
So all the software that CPU matters for?
kayk1@reddit
If you ignore all the stuff that doesn't work, then it's golden!
Immediate_Fig_9405@reddit
could still be better than x86 laptops that heat to 500C for running notepad.
marco_il_bello@reddit
Qualcomm Ceo will open Computex. Did someone one from 120$ to 250$ in one month or less? and contract with bytedance to sell millions of Ai chip ?
DehydratedButTired@reddit
I can’t wait for the performance metric they invent to be the best at.
the_dude_that_faps@reddit
It'll hallucinate benchmarks using AI and that way it won't even need to run them.
dib1999@reddit
Finally my joke about DLSS Game Generation stops being a joke
BrowsOfSteel@reddit
They’ll call it “silicon sampling” like the political pollsters.
BogiMen@reddit
I think it will be full of AI bullshit metrics
randomkidlol@reddit
100 gigarays per second
FrogNoPants@reddit
They will declare their SIMD lane count as being their core count :)
got-trunks@reddit
I can't wait for them to be for lease only with a 6k+ buyout.
VenditatioDelendaEst@reddit
The metric is ARM-Windows driver correctness, I'm calling it now.
AmazingSugar1@reddit
What’s the point of 6144 cuda cores if you can’t run x86 apps (games?) that’s a lot local AI processing
Moscato359@reddit
X86 emulators exist
That's how apple switched to arm
They made an x86 emulator on arm, and then started forcing all the software companies to start making arm versions
pythonic_dude@reddit
For games, switch to Linux and use fex+proton. For productivity, ms will have to deliver (and disappoint, really).
billiebuster@reddit
We all know that the ”new era of PC” that they want is cloud computing subscriptions because datacenters pay better
From-UoM@reddit
TIL - Client APUs for Laptops = cloud computing
imKaku@reddit
We used to call those thin clients.
From-UoM@reddit
Well there are two variants. The N1 looks to be one for thin and lights.
N1X is the GB10 chip with a rtx 5070 core count gpu in it. So it's a extremely absurd that people are cloud related.
Or even more bizarrely they probably think you need a rtx 5070 class gpu for for cloud gaming.
PJBuzz@reddit
Depends how much VRAM it has.
If the whole purpose of the GPU is edge processing, i.e. you upscale a shitty 720p stream, then you need a reasonable amount of GPU/NPU cores, not so much VRAM.
How much is on the consumer devices will tell us what they think about us owning our own hardware.
Serprotease@reddit
It’s teased to be an alternative to the amd ai max 395+ and a laptop version of the gb10. So, 32gb for the N1 and 64,128gb for the nx1 are likely (unified ram/vram, with 75% that can be allocated to the gpu on windows, or total -8gb on Linux.)
ResponsibleJudge3172@reddit
It's an APU. The RAM is the VRAM
bogglingsnog@reddit
I don't even need a 5070 in my main gaming desktop for 4k gaming, I have no idea why I would want a laptop with one.
OwlProper1145@reddit
Efficiency. Lots of CUDA cores running at a low clock speed is more efficient than a small amount at a high clock speed.
bogglingsnog@reddit
So basically, it matters when you are trying to game on battery? Relatively rare use case but okay that's a valid point.
BobCFC@reddit
it stops the fans spinning up as well
bogglingsnog@reddit
I suppose, but it's also basically a non issue on a well designed laptop with a proper cooling solution.
Die4Ever@reddit
Well it certainly doesn't stop them, but it means less cooling required for the same performance
Wetzilla@reddit
It's good for anything that heavily uses the CPU. This is an ARM SoC chip, not just a GPU.
Zarmazarma@reddit
So I have a 4090, which is about 70% faster 5070, and I'd say it's just about what you need for 4k on desktop. I think you are uniquely tolerant to low settings/low FPS if a 5070 is all you need for it. I mean I think it's a valid way to game, but I don't think most people would be satisfied with the 5070 @ 4k experience.
bogglingsnog@reddit
I use a 3070 and I have a 144hz screen and it's rare to come across a game so dog shit optimized that I can't get at least 100fps on it.
Zarmazarma@reddit
This isn't a thin client though. It's an ARM APU with a 5070~ tier GPU. The Legion 7 model takes 245w... it's not even like they're low power devices or anything.
From-UoM@reddit
There is two chips.
The N1X is the bigger 5070 one.
The N1 is going to be on thin devices like the Legion Ideapad Slim.
iad82lasi23syx@reddit
This is the weirdest conspiracy gamers have been pushing, there's nothing that points in that direction
BuldozerX@reddit
Because Gamers Nexus is their god.
billiebuster@reddit
I have never watched gamers nexus
Chimbondaowns@reddit
They make good quality content. You should.
Extreme-Arm4609@reddit
But not anymore man after that whole forced Linus drama they tried to do where they tried to basically start a civil war in that community and it never really recovered from there they've just been downhill.
Crappy rage bait content I don't like them anymore I don't watch them anymore I don't have a reason to I don't want to watch something just so negative and wrong if I want to watch something negative I'm just going to read what's going on in the industry if I want to sulk for a little bit
SituationSoap@reddit
They used to.
hsien88@reddit
you read a tweet from someone who watched gamers nexus. you are part of the fake news campaign and not even knowing it.
Devatator_@reddit
I really don't see why so many people still watch this clown
EdliA@reddit
Because people love drama and he only makes drama bait content
ResponsibleQuiet6611@reddit
Not sure what GN you guys are watching but it's not gamers nexus.
ResponsibleJudge3172@reddit
Go look at the last 30 videos and tell me more than half are not drama
DerpSenpai@reddit
I'm not sure why Gamers Nexus is even allowed on this sub
Avsunra@reddit
GN is good at tech and gaming, but that doesn't mean they know shit about economics or businesses at the scale of operating a data center. Nothing wrong with listening to their tech and gaming related content, but the moment they get into fields that really have nothing to do with their specialty I stop paying attention.
This isn't unique to GN, I follow content creators for all kinds of topics and people often say dumb shit the moment they stop talking about their own specialty. I encourage everyone to read and learn about various topics about the world, use your brain. Learn about economics, finance, business, geopolitics, maritime shipping and logistics, defense, international relations. The more you learn the more you will realize that most people don't know what the fuck they are talking about unless it's their specific specialty.
SlightComposer4074@reddit
His schizophrenic meltdown over the past couple years would've been funny to watch if he hadn't taken like 30% of online PC gamers with him
BuldozerX@reddit
The conspiracy is that Nvidia along with everyone else is teaming up to make hardware too expensive.
soggybiscuit93@reddit
It's not really a conspiracy. The demand for compute and memory currently exceeds the world's capacity to build it. And when expensive datacenter components are built from fundamentally the same capacity constrained facilities as cheap consumer stuff, then companies are going to prioritize the customers buying the super expensive high end components rather than people buying the lower cost stuff.
Everything we're experiencing now is the result of a rapid demand spike outstripping the ability for suppliers to keep pace. Companies may be shifting their businesses practices in response to this, but they're not spending $trillions to get people into DaaS.
BuldozerX@reddit
The data centers are not being built for cloud gaming, so the infrastructure isn't there, and the internet infrastructure also isn't there. It doesn't really matter because these cloud gaming subscriptions are going to be expensive as hell before reaching mass adaption, which would basically replace expensive hardware with expensive subscriptions.
soggybiscuit93@reddit
The same cost increases that consumers face for buying dGPUs are also impacting cloud gaming companies, and those price increases for their own procurement will be reflected in their subscription pricing.
It's not like they're paying regular price and consumers are paying exorbitant prices. Everybody is paying more, so the people doing most of the buying are those who believe that they can make a return on that hardware - and that's hardly cloud gaming services. Nvidia even introduced rate limits on their service, likely because they've slowed down expanding the hardware for the service because it's a waste when they could redirect that hardware to meeting AI demand.
VenditatioDelendaEst@reddit
Obviously the notion of a capitalist plot to destroy the desktop PC is ludicrous, but in fairness, Geforce Now does enable more hours of gaming per GPU. In relative terms the statistically multiplexed GPU goes up in price just as much as the privately owned one, but in absolute terms you're spreading the price increase across ~10 (?) gamers.
You should expect some of the gaming GPU market to shift from private ownership to streaming as the price goes up.
BuldozerX@reddit
It's obvious that the subscription price will increase sooner rather than later.
VenditatioDelendaEst@reddit
AFAICT, the price has been static since they introduced the *80 tier at $20/month in 2021. They adjusted the non-US prices for inflation in 2023, but OTOH, they added 120 FPS support.
But, supposing it does increase... so what? If at any point it becomes more expensive than owning a local dGPU, just buy one and cancel.
qtx@reddit
Hardware isn't even that expensive. It's the young'uns that think it's expensive, it isn't. It's cheaper (even these days) to buy a computer than it was in any of the previous decades.
maximalusdenandre@reddit
The dumbest part is that GeForce NOW already exists. The "rent a gpu" era is already here. You can go right now and get a subscription for cloud gaming on a computer with a 4080 (last time I checked). Why would they announce it now? And with Microsoft that runs a rivaling cloud gaming service in well, Xbox.
machinador@reddit
hate generate views and karma
Wetzilla@reddit
You can get a 5080 on a lot of games now.
G-Geef@reddit
The AI data center GPU's can't even run games anyways, it doesn't make sense
pigeieio@reddit
It's not something they may be even planning, it's just the natural development of data centers eating all of the resources and getting all of the focus from all of the parts makers. It's easier for them to just sell you some open compute time then maintain a less profitable and predictable consumer market.
Seanspeed@reddit
This subject is literally about them selling physical PC hardware. :/
jsheard@reddit
Well, even so, the fact that I could believe it tells you something about the current state of affairs
your_mind_aches@reddit
"Ah, well, nevertheless."
BuldozerX@reddit
Stop with this BS. Just because someone made a YouTube about it doesn't mean it's real. The price increase is due to AI data centers, and they are not being built to host cloud gaming. Nvidia improving DLSS, drivers and pushing out new chips also suggest otherwise. We also don't have the infrastructure to support this conspiracy theory.
constantlymat@reddit
The mass hysteria surrounding the supposed assault on personal computing is mindboggling to me.
As recently as late last fall, building a powerful 1440p gaming computer has never been more affordable.
A 9070XT for $600, 32GB of RAM $79, 2TB PCIe 4.0 SSD $100, a good B650 board $140 and a Ryzen 7800X3D a mere $330.
I cannot remember a better time to build a PC as powerful as this in my lifetime.
This has been just 9 months ago. People behave like this is the first really bad pricing bubble. Their accusations of a paradigm shift towards cloud computing has barely any supporting evidence.
cadaada@reddit
Where are you getting the ram and ssd for that price?
qtx@reddit
Re-read the comment, slowly. They're not talking about now.
cadaada@reddit
Yes i hate english thank you
NeroClaudius199907@reddit
Amd/intel doesn't register when they're creating outrage
DerpSenpai@reddit
It's because they cannot afford the most expensive graphics card Nvidia offers and so they paddel this BS. Nvidia since 2016-2017 and AI has become a thing, they have started pushing their highest end consumer cards for that use case. I worked on the same office space in 2018 as one of the earlier labs in my country using AI for voices for example (that would be deployed in SOTA text to speech). Every employee had a 2080 ti on their PC.
Deep90@reddit
This cloud pc narrative has no teeth.
These tech companies would need to prop up the price of hardware in perpetuity which is not sustainable.
Especially as the prices themselves have made it more viable to launch competitors and increased production in the coming years.
Soggy_Association491@reddit
Cloud PC needs some killer games to reach mass adaptation. May be some sort of RPG that cannot be datamined, recreating the wonder of exploration in mmorpg in the early 90s-2000s
BuldozerX@reddit
And they will have to increase the price of the cloud gaming subscription to a point where it's not considered cheap from our point of view.
From-UoM@reddit
Cloud gaming will become like cloud storage.
Not a replacement but a good to have extra option.
BuldozerX@reddit
Not become. It's already been there since OnLive. Thats 17 years ago. They have failed one after another. Geforce Now is basically the only thing that's left. Maybe xcloud.
MoreFeeYouS@reddit
Onlive irritated me with an input lag. GeForce Now irritates me 17 years later for the exact same reason.
You cannot break the speed of light.
Iron_on_reddit@reddit
Not with that attitude! We just need to discover wormholes, that's all.
RHINO_Mk_II@reddit
Stadia did, that's why they advertised "negative latency!"
semidegenerate@reddit
They just need to use AI to predict our actions. How hard could it be to model every individual customer's brain?
Wetzilla@reddit
I use geforce now regularly and the lag is very minimal for me.
zeronic@reddit
Likely depends on your local infrastructure and how close you are physically to their datacenter. The closer you are, the better your experience will be.
And given the state of telecom in general in the US and many parts of the world, it just doesn't make sense for many people if latency is a concern.
cheeseybacon11@reddit
Luna or Shadow?
soggybiscuit93@reddit
There's just so many issues with this conspiracy. One, most people are confused by the idea of enabling OneDrive sync on their PC's to sync some folders with the cloud. To think that those same people are going to even understand the concept of remotely controlling a cloud PC, or where to even go to begin signing up for one of those services is absurd. The average person doesn't even know what AWS or Azure is. The PC would have to literally hold their hand as part of the OOtB experience and boot them right into a cloud desktop....
...And as you've talked about with pricing. The increasing prices of RAM also increases the price of the hardware running these cloud PCs. What're they gonna do, buy a bunch of overpriced servers and hope to make the money back by getting grandma to use a hosted VDI service for $10/month? Genuinely makes no sense. The ROI is just simply not there.
What'll likely happen (and already is) is that RAM and storage in new PC's is going to stagnate and even go down, and the market will have to respond to less memory being available to the average consumer.
BuldozerX@reddit
Geforce Now subscribers will become slow boiling frogs soon enough. These services won't remain cheap forever, and they have already added 100 hour limit parental control.
AnonsAnonAnonagain@reddit
Especially when that capacity could be used hosting AI models
upbeatchief@reddit
And i am willing to be take a "lose" on electricity bills to enjoy my gaming pc. Your average data center will need to see returns for the chips, storage, memory, electricity, data center construction and local local taxes.
There's no world where a company can have enough VC funding to try to tackle cloud compute on such a massive scale.
Aggravating-Dot132@reddit
For your info, it is ALREADY A THING. There are houses in US that work as a router for the data center and they rent their hardware for people there. As well as electricity.
VenditatioDelendaEst@reddit
You need to spend less time on Bluesky.
LinkesAuge@reddit
Datacenters don't pay better. GPUa always had and still have very good margins and unlike giant corporations that spent billions on datacenters gamers don't have leverage to squeeze prices on individual units.
The difference is really just down to scale. Gaming (consumer GPUs) simply don't have the same upper bound as AI is now showing so you serve/prioritize the customer base that you can sell most too even if the individual margins are smaller.
Besides that I think people are overreacting to a short span in time and the current market situation.
This won't last forever and while in the here and now the AI crunch has a lot of downsides for gamers lets not ignore what this means for tech/GPUs in the long run and that is simply massive investments which will eventually result in much quicker progress and thus better prices for performance.
The leaps in AI performance for hardware are already gigantic between generations and while this obviously won't translate to gaming related hard performance in the short to medium term it is literally impossible for it not to translate to it sooner or later in a massive way (be it direct or indirect).
ghenriks@reddit
Nvidia wants many different things and making inroads into the PC market has potential to be a very good market to be in when the inevitable AI sales drop off
HellsNoot@reddit
The reality of it is that cloud GPU's can be utilized close to 100%, while consumer computing might be doing nothing for long stretches of time. Local will always be much more expensive than cloud computing.
account312@reddit
Do they have enough higher utilization that if everyone moved all their stuff to cloud, Nvidia wouldn't sell as much hardware?
cig-nature@reddit
Yep, smaller GPUs with less memory for more money is what I'm expecting.
LePfeiff@reddit
Please actually read the linked article before commenting :)
Lost_Tumbleweed_5669@reddit
I will go offgrid before I submit to the cuckoldry that is cloud computing.
SpitneyBearz@reddit
omfg!
shovelpile@reddit
I don't see how cloud gaming datacenters would pay better than consumers. Datacenters care way more about hard specs and price/performance ratios and they employ professional negotiators who expect high volume discounts.
BuldozerX@reddit
They have never literally said that they want 100% of gaming in the cloud. They have said thst they want people to have their own AI client. They were not talking about gaming at all.
DavidsakuKuze@reddit
No thank you. ARM vendor lock in cancer can fuck right off back to phones.
Extreme-Arm4609@reddit
What is even your point?
X86 is more locked down to literally 2 companies.
Arm is a lot less locked down vs that.
wh33t@reddit
+1, modular ISA or bust.
From-UoM@reddit
I am more interested in the N1 chip. It sounds like the full chip is the N1X.
Is the N1 a cut down GB10 or a new low power chip?
Also looks like every single major laptop OEM is going to have this chip.
alabasterskim@reddit
I'm SO glad to see a Surface rumored with N1X. Hopefully it's a Surface Laptop Studio 3, explaining away the fact they discontinued the SLS2 last summer + didn't release one yet this year since they were waiting on these chips which have kept getting pushed back.
JahEthBur@reddit
Man, purchased a Surface and it was honestly the worst laptop ive ever encountered.
I had to set it in the sunlight or hit it with a heat gun for it to even boot.
alabasterskim@reddit
Sounds like you got a dud, unlucky.
JahEthBur@reddit
Most def. Mad me sad inside and started me down the "nope" to MS path.
Its a hard road sometimes.
zxyzyxz@reddit
Bring back the detachable one, Book 4 please
Seanspeed@reddit
Nobody knows.
DerpSenpai@reddit
If they can sell me a 2k€ (tight budget) Nvidia N1 im sold else im going for the A16 with the X2 Elite. I can forgive the "only" Panther Lake like performance in late 2026 if it's a good product (good battery life, GPU delivers)
Idkwatnam@reddit
They could've improved the performance a lot since the result I am bringing up but the CPU in the N1 according to early Geekbench results is more like Panther Lake "U" performance / around the same as a Snapdragon X Plus:
alps evb_8091_la - Geekbench
Idkwatnam@reddit
The N1 should be a cut down version of the GB10. Geekbench results show it as having 10 cores (5P +5E), so basically one of the two clusters of the full GB10 / N1X and as many CUDA cores as an RTX 5050 instead of a 5070. Presumably the memory bandwidth would also get cut in half but I don't think there's a way to check that from Geekbench results.
JahEthBur@reddit
Lowkey feeling like Nvidia is dead to for what they did to my home PC bottom line.
Next GPU will likely be anything other than a Nvidia card.
__some__guy@reddit
New era of Personal Clouduting.
nonaveris@reddit
You will own nothing and be happy.
soggybiscuit93@reddit
Setting aside whether Nvidia actually wants to commit the necessary effort or not, imo they're probably the only current ARM SoC maker on the market that has the market-command and brand recognition to really move the WoA adoption needle in ways that Qualcomm could only dream of.
MikhailT@reddit
Why wouldn’t they want to, it’s extra revenue on top of their custom arm design works for their server designs; just downscaled for consumers.
Expand their CUDA / software/ AI ecosystem to this and they will have more locked in customers for a long time since other companies will have a hard time catching up.
Vince789@reddit
A reused server design is why I'm not really interested
The N1X is clearly a server chip first & foremost, with its weird 10P+10E CPU, 5+5 clusters, asymmetric cache, ...
I highly doubt the N1X is going to standout in either efficiency or value, and its probably gonna be low volume too
VenditatioDelendaEst@reddit
Rather the opposite, I think.
Hybrid designs are bad for server, because either you affine to P-cores and leave the machine more than half idle, or 99%+ latencies will be dominated by E-cores. You might have a workload that's not CPU-latency-bound, but then you'd get more perf/$ from a chip that was all E-cores.
Vince789@reddit
Agreed, that's what I mean by I'm not interested in a reused server design
I'd rather see pure consumer & pure server designs
Nvidia does have pure server designs: Grace & Vera CPUs
But no pure consumer design yet
This N1X/N1's CPU is designed by MediaTek, and its a weird hybrid server design
VenditatioDelendaEst@reddit
Why are you saying "agreed"? You said the hybrid architecture indicates it's a server-first design. I said hybrid is bad for server (implication: hybrid indicates consumer-first).
You say you agree and reiterate that it's re-used server?
IMO, the cluster including both P and E cores is something you'd only do for client. Look at AMD. The server/desktop chiplets are all Zen or Zen Dense. The mobile parts are hybrid.
What I was able to Google up suggests that the CPU part of GB10, and presumably N1X, was designed in partnership with MediaTek, whose lineage is mobile client SoCs. Which have been using multiple types of cores and clusters since before Intel.
AFAICT there is nothing server-y about N1X.
Vince789@reddit
The confusion steems from the comment I initially replied to, which said the N1X/N1 is Nvidia reusing their server designs, I was too lazy to correct that
N1X/N1 is Nvidia reusing their GB10 Superchip from the Nvidia DGX Spark, which is weird sorta hybrid server / workstation / mini-PC design (not pure server Grace/Vera CPU, or pure client)
For Client, you'd want P & E cores in separate cluster for scheduling & efficiency/power management. Just like with Apple, Qualcomm, AMD's client designs. P and E cores in the same cluster quite rare, seems almost like something they'd do for automotive, which is where some of their DGX chips also go
10P+10E is really weird for an Arm Cortex CPU, its too many E cores, similar to the Exynos 2600
Arm themselves recommends 12P+4E for laptops since the X4/A720/DSU-120 (no update since then)
VenditatioDelendaEst@reddit
Rumor is that N1X/N1 has been delayed so long because they had to fix/work around a shedload of bugs, of the sort that would make it non-viable as a mobile SoC.
My guess is that it was first productized in DGX Spark so that they could expose it to a user base in a situation where it wouldn't matter if it idled at 20 W or built-in peripherals didn't work and had to be substituted with 3rd party USB/PCIe ones.
Vince789@reddit
My guess is the DGX Spark was always their priority. They struggled to optimize their workstation design to be competitive with Apple & Qualcomm's laptop chips in terms of efficiency/battery life (excluding gaming)
Which isn't surprising, as explained previously, no other client chip has such a weird CPU design with obvious scheduling & power consumption issues
soggybiscuit93@reddit
Because they would have to devote a pretty serious amount of silicon to gain notable marketshare in laptop in an era where theyre struggling to keep up with much more lucrative datacenter demand.
ResponsibleJudge3172@reddit
That does not stop them from diverting some resources to car chips, which have been stuck at 100 million a quarter for 5 years, why the fear mongering does not start at those weak sectors first is beyond me
soggybiscuit93@reddit
Nvidia isnt going to let healthy markets theyre highly competitive in (or dominate in) just die. They're not going to abandon automotive or client just because datacenter is so profitable...
But at the same time, N1X is essentially a repurposed server CPU and theyre gonna wanna prioritize server sales over laptop. They would need to sell a huge amount of N1X to make WoA really grow to reasonable amounts. So they'll participate in these lower markets, but theyre not going to do so with aggressive volume.
Zarmazarma@reddit
Nvidia is forward thinking though. Having the foothold in this market means that they'll be ready to take over if it explodes, and AI demand falls off or something. I mean, Nvidia was developing the hardware for AI long before it exploded- it's why they're the ones making $130B/year now, because they had the shovels to sell for the gold rush.
LastChancellor@reddit
~~they didnt actually make the CPU, they hired Mediatek to make it for them~~
alabasterskim@reddit
Would be absolutely hilarious if they again don't announce at Computex.
siazdghw@reddit
People keep saying these chips are for an imminent release and then months go by and nothing.
According to news outlets and reddit Nvidia was already 'supposed' to launch these chips a few times already...
ResponsibleJudge3172@reddit
Arm, Nvidia and Microsoft are advertising the launch on their platforms
alabasterskim@reddit
Yup. I'll believe it when I see it. Based on the article it's definitely coming together, but I really wouldn't be surprised to get let down again. I have heard that at this point, if they delay again, they're gonna have to scrap the chip altogether and just launch the next iteration, which I imagine they don't want to do.
Dizman7@reddit
“All future gpus will be subscription based”
Swoop03@reddit
Free ad supported tier limits to 4Gb memory $19.99 per month to an ad supported 8Gb locked at 1080p 30fps $35.99 per month for ad free access to 12-16Gb for smooth 1440p gaming at 40 fps $44.99 monthly for our ultimate unlimited gaming access and a non removable AI assistant
Dizman7@reddit
Those prices are way to low. Highest tier will be like $200/mo
Equivalent-Money9756@reddit
Ah yes, the "we've decided to pursue a new framework and force you all to pay us for new hardware again to keep up" bit. I sorta have a hard time believing that ARM could overtake x86. It's just too inefficient for gaming, and doesn't have support for tons of specialty software. Hell, just my basic music DAW only supports ARM through emulation. DAWs are really high demand software, usually maxing the CPU on most computers once a project file hits its full size. The performance hit of running on ARM in certain situations alone is going to keep it niche im thinking.
TheEvilBlight@reddit
The real kicker is compat handling with a massive legacy support base. Thats the moat arm hasn’t been able to crush.
marios_geo2@reddit
It's really happy /sad times. We finally get powerful cpu/gpu with lots of memory (apple or nvidia order .. doesn't matter really) so we speed up our worflow with daily hardware.
Yet the prices have sky rocket. Fuck
TheEvilBlight@reddit
Maybe they’ll follow the Apple integrated memory approach and cut out the memory dealers. But storage prices, Augh
TheEvilBlight@reddit
“Based on our SoC experience with the switch2, behold our NVARM chips!”
Allfeelings0Logic@reddit
Remember guys Nvidia is done with consumer hardware, they only care about data centers now.
ResponsibleJudge3172@reddit
Gamer Saltiness channel exclusive leak
el_f3n1x187@reddit
Its soo going to come with a subscription to so,e e,beded cloud service
Simple-Sound4405@reddit
Is this recent information? I hadn't seen this before.
mombi@reddit
It's gonna be a cloud subscription for compute power isn't it
TESThrowSmile@reddit
Do you think he wears a new leather jacket everyday ?
jjason82@reddit
So tired of this clown and his stupid jackets.
AnechoidalChamber@reddit
New era of PC...
Terminal plugged remotely into their AI servers installed on your lawn or backyard that they own but uses your space, your water and your electricity that you pay and of course pay the subscription or no games/apps for you.
MaximumEnshittificationAccomplishmentUnlocked
jsheard@reddit
Yeah I'm sure they made a chip with 20 cores and as many GPU cores as a 5070 with the intention of it being used as a thin client.
AnechoidalChamber@reddit
The plans are already in motion for exactly what I described and Nvidia is the pillar of it all.
You need to see further than this announcement, what is described above is their end goal.
And excuse me, but I can't take a laptop seriously as anything other than a thin-like client. Laptops are more or less the Paralympics of PCs just by how crippled they are power and thermals wise, especially in an era where both have been steadily climbing ( 250W CPUs, 600W GPUs ), what is the highest power rated laptop at? In the hundred watts range?
RoyKami@reddit
You do realise that laptops aren't just used for gaming only? Also gaming laptops especially 50 class and 60 class have identical or very similar performance to that of desktop variants. They both fit within 100-110W of power profile only meant for the gpu.
Let's say they are trying to support this own nothing rent everything. Why will they make their laptop GPUs perform this good especially at lower ends where the biggest consumer base is located? Why would they Make the GPUs be this good that allows the user to play 97-98% of the games available play at max settings 1080p. Make your comment make sense mate.
Look beyond the fear mongering bro, cloud gaming is just an extra for those that have very fast and good internet and a huge part of the world doesn't have that.
Morgenlatte29@reddit
Practically every point in your comment is absolute nonsense lol
AnechoidalChamber@reddit
If that is your truth, I take it then your truth is also that war is peace, freedom is slavery and ignorance is strength.
Good luck with that.
Morgenlatte29@reddit
What. Lol
doscomputer@reddit
Well when it has half the bandwidth of the 5070 and has to share all of that overhead with the CPU, its definitely going to perform like a thin client.
VenditatioDelendaEst@reddit
Where do you people get this shit? What is the server going to do with the water? Where does it go?
akanosora@reddit
It electrifies water into hydrogen and cold fusion hydrogen to generate power for its AI cores.
ConsistencyWelder@reddit
Can't wait for everyone not to buy one.
Latrodectus1990@reddit
I would never ever pay for cloud gaming
Fu*k Off with that crap
doscomputer@reddit
downvoted for speaking truth
996forever@reddit
Downvotes for being off topic but that has nothing to do with this big apu with 20 cpu cores and same amount of gpu cores as a 5070.
Mr_Resident@reddit
nice new device with vibe code driver
Exciting-Possible773@reddit
Can it run games? Because GB10 can't.
ResponsibleJudge3172@reddit
Interesting that Mediatek is not visible as the CPU partner
charliehu1226@reddit
They are partnering with Google’s new AI PC.
Luggage-Lock@reddit
It would negate some of the hype from having Nvidia branding on there.
Vince789@reddit
Yep, Mediatek still a bad rep. They some great chips, particularly below the flagship tier
But its still somewhat deserved since they are clearly behind Qualcomm, Samsung & Xiaomi in implementation at the flagship tier
sn2006gy@reddit
What is the pricepoint on these? anyone know? The Sparks are at 4-5k and feel like a rip off - are we expecting these to be 3500 bucks or more too or will they disrupt the spark pricing because these could be ultra portal dev machines for CUDA?
Luggage-Lock@reddit
Sparks are going for 6-8k last I checked. These will likely be well over the 3500 mark. Nvidia is trying to get some massive margins from the OEMs on these to capitalize on the company’s hype.
I honestly don’t see this form factor selling as well as the DGX Spark, especially with how power hungry it is.
sn2006gy@reddit
I bought my Asus GX10 Spark for 3.5k and it's not really power hungry and no way these could be over $3500 mark
Luggage-Lock@reddit
Guessing you bought before the new year. Prices have exploded the last couple of months. Seeing these quoted regularly in that 6-8k range, even in volume.
doscomputer@reddit
3 upvotes on a completely and utter lie? Nvidia marketing is here on this post arent they?
wouldn't want people buying the $3500 ASUS model when the official one is $6500 from nvidia right?
sn2006gy@reddit
the OEM version went up in price, but the 3rd parties can be found for 3.5k a piece on amazon and other places still. Asus doesn't seem to be raising prices to match Nvidia's.
Idkwatnam@reddit
I expect them to be less than that yeah. Remember the Sparks come with a Connect X7 NIC which individually cost around $1500 USD. The Sparks also come with 128GB RAM only, drop the RAM to maybe as little as 48GB and devices with an N1X could be a LOT cheaper than the DGX Sparks.
doscomputer@reddit
A new era of E-waste and electronics build to a minimum price specifically to be useless after so many years
airfryerfuntime@reddit
Can we have a new era of affordability?
77ilham77@reddit
"Best we can do is renting"
Darksider123@reddit
Era of laptops and "personal AI" he called it. Our hobby of powerful dedicated gaming desktops are coming to an end according to him
RedditNotFreeSpeech@reddit
"you can rent one from us!"
Kougar@reddit
The tone deafness of heralding a new era for PC when your other products are to blame for making PCs unaffordable. Can't supply enough GPUs to the consumer market, but can find the silicon to supply new CPUs to them just fine?
jedrider@reddit
Are these chips just to upend the Qualcomm ARM business but these chips are actually Risc-V architecture?
GenericUser1983@reddit
These chips are not Risc-V, they are ARM based cores. The NX1 chips uses 10 Cortex-X925 + 10 Cortex-A725, the N1 IIRC cuts the core count in half.
coffee_and_stims@reddit
i'm sick of seeing this rat faced fuck.
cloud_t@reddit
Translation: "The new era of PC is here, because we decided it for you".
Alexix1996@reddit
Un chip con una potenza di eGPU integrata pari alla 5070 desktop, mettiamoci dentro il frame generation evoluto nel DLSS5...con circa 200W sarà possibile giocare a praticamente tutto a Ultra a 4K a FPS molto alti. Staremo a vedere NVIDA cosa sforna a questo giro.
GroundedGeeking@reddit
For desktops , I'd never buy anything that didn't have simple to replace CPUs and GPUs. Laptops and SBCs are fair game but while I'm good using Tegra boards at work, in the end I expect someday to be using RISC-V machines
K33P4D@reddit
Gimme a 6060 16GB for 350$
Yojik_Vkarmane@reddit
Now you can have fake frames already in BIOS.
battler624@reddit
Just gotta wait for performance figures, if its anywhere close to their 5070ti mobile with high compatiblity then x86 could very well be in danger.
As afaik, this will be the first arm cpu with a well known gpu/driver stack.
SirMaster@reddit
Isn't the N1X like a year old at this point now?
verynotfun@reddit
I believe NOTHING from that guy
Banana-phone15@reddit
Shut up Jensen. You have lost your credibility and respect in PC market. 🖕
GhostsinGlass@reddit
This isn't PCMR.
imKaku@reddit
I’ll be surprised if it’s anything but a rebrand of their game streaming platform.
Seanspeed@reddit
Did you respond in the wrong comment section? This is literally about a new high end processor they're gonna release for laptop, with their own ARM cores and a huge GPU.
imKaku@reddit
My mistake!
Stilgar314@reddit
Sure that new era is some AI related shit
SpitneyBearz@reddit
Oh god, is he killing PCs? Laptop Chip?!?! "New era of PC"?!??!?!
kivimango23@reddit
We will call it a funeral.
Didgeridoo69420@reddit
I wouldn't be surprised if this eventually becomes a "NVIDIA Premium Thin Client" that's required for "Priority Access" to GeForce Now.
kittymoo67@reddit
you're gonna need to get actual backwards comparability into arm before it can replace x86 in the consume market. and also let us build our own.
wondersnickers@reddit
Fuck Them. Always artificial scarcity, bandwidth and ram locked behind prestige premium prices.
AutoModerator@reddit
Hello PaiDuck! Please double check that this submission is original reporting and is not an unverified rumor or repost that does not rise to the standards of /r/hardware. If this link is reporting on the work of another site/source or is an unverified rumor, please delete this submission. If this warning is in error, please report this comment and we will remove it.
I am a bot, and this action was performed automatically. Please contact the moderators of this subreddit if you have any questions or concerns.