Not sure there's a strong use case for one of these $6K laptops yet, considering you can get a mini PC running AMD Ryzen 495 or 395 with 96, 128 and 192 GB of memory for similar amount of money and connect remotely from a cheaper/smaller/more battery-efficient laptop. Microsoft also has this same chip in non-laptop form factor: https://www.microsoft.com/en-us/surface/devices/surface-rtx-....
Coding while flying is often the use case people bring up, but the power brick these laptops come with sometimes trip the plane's power plug in coach and you can't run local AI on battery for long. Also, over time planes will get better connectivity (Starlink etc.) and that particular full-offline scenario will go away.
For gaming, there are better options with dedicated GPUs like the Asus Zephyrus and similar.
The core design is also approaching 3 years old at this point, so it’s not really even that competitive with the latest stuff out of the other chip makers. It’s fine, but nothing to write home about. Certainly not worth the crazy premium Nvidia is charging for it.
The strong case is pretty much only if you can't wait another 3-4 years for ram prices to come back down.
Memory makers are responding with a huge build-out. Building a fab shell and cleanroom takes years, which is why the crunch has lasted so long, but many of these projects are now well underway. New capacity is what ended past memory shortages. New factories going in:
SK hynix
M15X, Cheongju, Korea: first wafers early 2026 (reported), ramping through 2027
Yongin Y1, Korea: first cleanroom Feb 2027 (company target)
Yongin Y2, Korea: groundbreaking planned July 2027, first cleanroom June 2029
Micron
ID1, Boise: first wafer output mid-2027
Tongluo, Taiwan: an existing PSMC cleanroom it bought for $1.8B, with meaningful shipments mid-2027
ID2, Boise: late 2028
Hiroshima, Japan: broke ground July 2026, with initial output late 2028
Clay, New York (up to 4 fabs, $100B): broke ground Jan 2026, with first output in 2030
Samsung
P5, Pyeongtaek: mass production 2028 (reported)
Yongin national complex: site work began Oct 2026, with the first fab targeting Oct 2029 (reported; Samsung hasn't given a date)
Nanya
Fab 5A, New Taipei: 10nm-class EUV DRAM, with wafer starts in 2H 2027 (reported)
Tesla / SpaceX (Terafab)
Research fab at Giga Texas, Austin: broke ground April 2026. There's no official output date.
Grimes County, Texas (logic, memory, packaging; Phase 1 over $16.8B): land clearing since Aug 2026. Texas filings say commercial operations in 2028, but SemiAnalysis estimates late 2029.
The first real supply relief is 2027, a second wave comes in 2028–29, and a lot of that new capacity is aimed at HBM and server memory, so consumer DDR5 and GDDR will likely ease last.
Sources: Micron FQ4 2026 call, the SK hynix newsroom, TrendForce, Texas JETI filings via ENR, and Tom's Hardware/SemiAnalysis.
And maybe 2027 is a truly optimistic relief on some level,
> a second wave comes in 2028–29, and a lot of that new capacity is aimed at HBM and server memory, so consumer DDR5 and GDDR will likely ease last.
'Consumer DDR5' has a pretty big set of knock-on effects though.
Unless you are working at a 'tech' company, the norm is that even developer machines will be constrained on memory usage for a given workload[0][1]. IOW companies trying to keep procurement in check wind up looking at procurement cost vs lost productivity.
[0] - Modern infosec and corporate software, at least where I'm at, seems to be vaguely more efficient than electron apps, IOW I can reboot, open teams and my IDE, after 4GBish of my IDE and 1GB of a fresh Teams instance I'm already around 16 of 32GB.
[1] - I had to fight hard to get 32GB as a developer baseline at my org. They wanted 16GB. Ironically the JVM based devs get 64GB but still have slowdown issues on biuld...
The Nvidia based one uses the memory for the GPU allowing you to run AI models on the GPU, the system you're talking about has no such setup. These are Nvidia powered laptops that do what Apple does with their laptops where the system and GPU VRAM are shared as needed, except the idea here is that Nvidia designed these to perform very well for local inference, at least, that's how I understand it.
Depending on what you are doing you may find all the models are written for CUDA though. I'm playing with gaussian splatting and it's like that. Codex is doing a good job porting stuff to Metal, but it would be easier & faster if I had Nvidia hardware to just run the models natively.
Not everything with AI is LLMs running on llama.cpp.
The AMD hardware is x86_64, which means lower battery life on average for normal users, and as an architecture, it is very well supported across the Windows ecosystem.
The RTX Spark is ARM64, which means more efficient power management at the cost of a small hit to sustained performance loads. But comes at the cost of being less supported when it comes to available software packages.
Aside from this, I’m led to believe that the AMD stack when it comes to local models has more overhead related to its need to translate more NVIDIA native instructions and therefore can be less performant than Nvidia.
Not sure how true this is anymore, but ive yet to hear otherwise.
> The AMD hardware is x86_64, which means lower battery life on average for normal users [...]
A Ryzen AI Max+ 395 powered tablet with 70Wh [0] gets the same battery life given the same task as a Nvidia N1X powered laptop with 99.9 Wh [1].
Power Consumption in their "Cyberpunk / Stress Test" is also lower at Ø109.3 (108.8-110.8) Watt for the x86_64 vs Ø134.5 (132.1-137.5) Watt for the arm64 system using the same settings, resolution and display brightness. They also get the same performance when normalising for resolutions and settings in Cyberpunk, so this is not an instance of one providing significantly more performance.
To add, battery life on this chip hasn’t been that impressive [1]. It’s basically matching recent MacBook Pros, while sporting a significantly larger battery. Both AMD and Intel have made significant strides in improving power consumption on their latest core designs, to the point where there isn’t a lot of difference anymore.
Why? Flying business everywhere is substantially more expensive than a one time cost of $6k for a laptop. A single business ticket anywhere overseas matches or exceeds that in most cases.
What? I have spent $10k on a computer but would never fly business or first class, that shit is way too expensive. Two trips would buy a new $10k computer!
Only 30 tok/s on Qwen 3.8 Flash Next? These days thanks to those specialized engines out there I can get ~2x that speed on a 24 GB video card with RAM offload. Thermal limits, I assume...
Being forced to login to use your own hardware is one issue for sure…
but separately everything about Microsoft’s login experience is pure pain.
Multiple broken redirects, broken links, buttons that loop, and more just to hit a CAPTCHA AI can solve anyway.
The only thing worse is for the poor soul developers who have had to login to their dev portal to setup SSO Microsoft for an app. It’s the worst I have ever seen in my life (even using Supabase that hand holds you through it).
I have a DGX Spark (Ubuntu), and when Microsoft and NVIDIA announced the RTX Spark running Windows, I thought to myself: Why would I get that instead of the DGX Spark?
The OS makes a big difference for most developers, and the DGX Spark stack and experience is mature.
I thought so too, but there's a ton of people out there wading their way through trying to re-purpose their gaming rigs to run qwen3.8-flash-next on their gaming rig (at Q2..) via WSL
> When rumours started circulating about a laptop with over 100 GiB of RAM the GPU could access
That was like three years ago with the release of M3, do you people not count MacBooks as laptops or something? Not that you could run Linux on it back then, but you don't even mention the existence of Apple silicon at all. :/
As someone who was intrigued by these new machines but also has no interest in running Windows, I'm quite happy to learn it works. Though admittedly I was expecting to see DGX OS first.
It is getting progressively more difficult to set up a new Windows laptop without a cloud account. Currently it's a pretty esotaric command line prompt if you know the keybinding to get a terminal open.
Windows Pro > Work or School Account > Join a Domain instead > Set password & recovery questions
The one component here that is not present on most machines straight from the manufacturer is the presence of the Pro version.
The home version of windows requires the account with increasingly limited alternatives.
As the author mentions in TFA, she's on the Arch staff as a package maintainer. I don't think there was a technical obstacle around the windows setup process.
Microsoft turns on OneDrive backups by default without your knowledge or consent, sending your private files to MS servers, not only during setup but it can also happen during updates. Even when you try keep it disabled, they constantly pressure you into enabling it every time you login, and make it really easy to accidentally turn it on. Once your files are in OneDrive, MS can access them. Don't be surprised if it later turns out they're training their LLMs on people's private files.
I don't want a cloud login telemetry of when I am using my computer, an account that can be disabled by a provider, and constant ads to backup to onedrive. I also moved to linux because of this. (was already a linux guy, but Windows is completely in my rear-view now.)
Not sure there's a strong use case for one of these $6K laptops yet, considering you can get a mini PC running AMD Ryzen 495 or 395 with 96, 128 and 192 GB of memory for similar amount of money and connect remotely from a cheaper/smaller/more battery-efficient laptop. Microsoft also has this same chip in non-laptop form factor: https://www.microsoft.com/en-us/surface/devices/surface-rtx-....
Coding while flying is often the use case people bring up, but the power brick these laptops come with sometimes trip the plane's power plug in coach and you can't run local AI on battery for long. Also, over time planes will get better connectivity (Starlink etc.) and that particular full-offline scenario will go away.
For gaming, there are better options with dedicated GPUs like the Asus Zephyrus and similar.
What am I missing?
The core design is also approaching 3 years old at this point, so it’s not really even that competitive with the latest stuff out of the other chip makers. It’s fine, but nothing to write home about. Certainly not worth the crazy premium Nvidia is charging for it.
Yeah but this is all there is until what 2028? 2029? And those could cost 2x-3x too.
May always be behind with how fast things change, you think?
The strong case is pretty much only if you can't wait another 3-4 years for ram prices to come back down.
Memory makers are responding with a huge build-out. Building a fab shell and cleanroom takes years, which is why the crunch has lasted so long, but many of these projects are now well underway. New capacity is what ended past memory shortages. New factories going in:
SK hynix
M15X, Cheongju, Korea: first wafers early 2026 (reported), ramping through 2027 Yongin Y1, Korea: first cleanroom Feb 2027 (company target) Yongin Y2, Korea: groundbreaking planned July 2027, first cleanroom June 2029 Micron
ID1, Boise: first wafer output mid-2027 Tongluo, Taiwan: an existing PSMC cleanroom it bought for $1.8B, with meaningful shipments mid-2027 ID2, Boise: late 2028 Hiroshima, Japan: broke ground July 2026, with initial output late 2028 Clay, New York (up to 4 fabs, $100B): broke ground Jan 2026, with first output in 2030 Samsung
P5, Pyeongtaek: mass production 2028 (reported) Yongin national complex: site work began Oct 2026, with the first fab targeting Oct 2029 (reported; Samsung hasn't given a date) Nanya
Fab 5A, New Taipei: 10nm-class EUV DRAM, with wafer starts in 2H 2027 (reported) Tesla / SpaceX (Terafab)
Research fab at Giga Texas, Austin: broke ground April 2026. There's no official output date. Grimes County, Texas (logic, memory, packaging; Phase 1 over $16.8B): land clearing since Aug 2026. Texas filings say commercial operations in 2028, but SemiAnalysis estimates late 2029.
The first real supply relief is 2027, a second wave comes in 2028–29, and a lot of that new capacity is aimed at HBM and server memory, so consumer DDR5 and GDDR will likely ease last.
Sources: Micron FQ4 2026 call, the SK hynix newsroom, TrendForce, Texas JETI filings via ENR, and Tom's Hardware/SemiAnalysis.
> which is why the crunch has lasted so long,
It's been like a year-ish...
And maybe 2027 is a truly optimistic relief on some level,
> a second wave comes in 2028–29, and a lot of that new capacity is aimed at HBM and server memory, so consumer DDR5 and GDDR will likely ease last.
'Consumer DDR5' has a pretty big set of knock-on effects though.
Unless you are working at a 'tech' company, the norm is that even developer machines will be constrained on memory usage for a given workload[0][1]. IOW companies trying to keep procurement in check wind up looking at procurement cost vs lost productivity.
[0] - Modern infosec and corporate software, at least where I'm at, seems to be vaguely more efficient than electron apps, IOW I can reboot, open teams and my IDE, after 4GBish of my IDE and 1GB of a fresh Teams instance I'm already around 16 of 32GB.
[1] - I had to fight hard to get 32GB as a developer baseline at my org. They wanted 16GB. Ironically the JVM based devs get 64GB but still have slowdown issues on biuld...
The Nvidia based one uses the memory for the GPU allowing you to run AI models on the GPU, the system you're talking about has no such setup. These are Nvidia powered laptops that do what Apple does with their laptops where the system and GPU VRAM are shared as needed, except the idea here is that Nvidia designed these to perform very well for local inference, at least, that's how I understand it.
No, the AMD ones he’s talking about allow the GPU to address that memory as well: they’re “APU’s” that share system RAM as GPU memory
That said the compute power of the GB10 is awesome, my DGX Spark-like has been better than I expected.
Depending on what you are doing you may find all the models are written for CUDA though. I'm playing with gaussian splatting and it's like that. Codex is doing a good job porting stuff to Metal, but it would be easier & faster if I had Nvidia hardware to just run the models natively.
Not everything with AI is LLMs running on llama.cpp.
That's the thing, my understanding is that these Nvidia ones are supposed to have very impressive performance.
The AMD hardware is x86_64, which means lower battery life on average for normal users, and as an architecture, it is very well supported across the Windows ecosystem. The RTX Spark is ARM64, which means more efficient power management at the cost of a small hit to sustained performance loads. But comes at the cost of being less supported when it comes to available software packages.
Aside from this, I’m led to believe that the AMD stack when it comes to local models has more overhead related to its need to translate more NVIDIA native instructions and therefore can be less performant than Nvidia. Not sure how true this is anymore, but ive yet to hear otherwise.
> The AMD hardware is x86_64, which means lower battery life on average for normal users [...]
A Ryzen AI Max+ 395 powered tablet with 70Wh [0] gets the same battery life given the same task as a Nvidia N1X powered laptop with 99.9 Wh [1].
Power Consumption in their "Cyberpunk / Stress Test" is also lower at Ø109.3 (108.8-110.8) Watt for the x86_64 vs Ø134.5 (132.1-137.5) Watt for the arm64 system using the same settings, resolution and display brightness. They also get the same performance when normalising for resolutions and settings in Cyberpunk, so this is not an instance of one providing significantly more performance.
[0] https://www.notebookcheck.net/Asus-ROG-Flow-Z13-GZ302EA-Conv...
[1] https://www.notebookcheck.net/Asus-ProArt-P16-RTX-Spark-Lapt...
To add, battery life on this chip hasn’t been that impressive [1]. It’s basically matching recent MacBook Pros, while sporting a significantly larger battery. Both AMD and Intel have made significant strides in improving power consumption on their latest core designs, to the point where there isn’t a lot of difference anymore.
[1] https://www.lttlabs.com/articles/2026/10/16/nvidia-rtx-spark...
A mini pc doesnt have a battery. Your point doesnt seem very coherent.
i don’t think they’re flying coach after spending 6k on a laptop…
Why? Flying business everywhere is substantially more expensive than a one time cost of $6k for a laptop. A single business ticket anywhere overseas matches or exceeds that in most cases.
What? I have spent $10k on a computer but would never fly business or first class, that shit is way too expensive. Two trips would buy a new $10k computer!
I got the HP version of this 128GB for 3.8k. I have yet to see a better deal for that ram.
Only 30 tok/s on Qwen 3.8 Flash Next? These days thanks to those specialized engines out there I can get ~2x that speed on a 24 GB video card with RAM offload. Thermal limits, I assume...
I wonder how its performance is on dense models.
memeory bandwidth is typically the limiting factor, not compute. Your card quite likely has 2x or more the bandwidth.
Being forced to login to use your own hardware is one issue for sure…
but separately everything about Microsoft’s login experience is pure pain.
Multiple broken redirects, broken links, buttons that loop, and more just to hit a CAPTCHA AI can solve anyway.
The only thing worse is for the poor soul developers who have had to login to their dev portal to setup SSO Microsoft for an app. It’s the worst I have ever seen in my life (even using Supabase that hand holds you through it).
I have a DGX Spark (Ubuntu), and when Microsoft and NVIDIA announced the RTX Spark running Windows, I thought to myself: Why would I get that instead of the DGX Spark?
The OS makes a big difference for most developers, and the DGX Spark stack and experience is mature.
I thought so too, but there's a ton of people out there wading their way through trying to re-purpose their gaming rigs to run qwen3.8-flash-next on their gaming rig (at Q2..) via WSL
> When rumours started circulating about a laptop with over 100 GiB of RAM the GPU could access
That was like three years ago with the release of M3, do you people not count MacBooks as laptops or something? Not that you could run Linux on it back then, but you don't even mention the existence of Apple silicon at all. :/
Hm, what's the learning here?
What challenge has been overcome?
I can see that hardware was bought, and that's cool hardware and all, but consumption is rather uninspiring.
I suppose it's the arm64 part? Or the review of some new hardware other people might be interested in?
She literally posted a link to the patches she wrote to get it to boot linux. That seems like it's not nothing
But does it go into detail about those changes?
Because that sounds interesting! But I didn't see anything like that when I've (admittedly) skimmed through the review.
As someone who was intrigued by these new machines but also has no interest in running Windows, I'm quite happy to learn it works. Though admittedly I was expecting to see DGX OS first.
i would do same if i have to use windows. I will literally pay anything to not use windows, thanks to microslop.
Is a Microsoft account required to use Windows on these RTX spark laptops?
Am I being too stubborn for not wanting a cloud account tied to my desktop operating system?
No, the vendor is being too presumptuous. This is clearly a case of them putting their needs above yours.
It is getting progressively more difficult to set up a new Windows laptop without a cloud account. Currently it's a pretty esotaric command line prompt if you know the keybinding to get a terminal open.
For Win10, I had to keep it disconnected from the network throughout the install process. Once you're connected, you have to do it all over again.
Win7 was the GOAT
Does anybody know how Wayland performs on these?
simply creating a Microsoft account would have been far quicker and easier ;-)
getting a local account is simple.
Windows Pro > Work or School Account > Join a Domain instead > Set password & recovery questions
The one component here that is not present on most machines straight from the manufacturer is the presence of the Pro version. The home version of windows requires the account with increasingly limited alternatives.
As the author mentions in TFA, she's on the Arch staff as a package maintainer. I don't think there was a technical obstacle around the windows setup process.
https://archlinux.org/people/package-maintainers/#josie
Simple™. Easy™.
Hey just as a heads-up Microsoft accounts are free to make https://signup.live.com/
Microsoft turns on OneDrive backups by default without your knowledge or consent, sending your private files to MS servers, not only during setup but it can also happen during updates. Even when you try keep it disabled, they constantly pressure you into enabling it every time you login, and make it really easy to accidentally turn it on. Once your files are in OneDrive, MS can access them. Don't be surprised if it later turns out they're training their LLMs on people's private files.
I don't want a cloud login telemetry of when I am using my computer, an account that can be disabled by a provider, and constant ads to backup to onedrive. I also moved to linux because of this. (was already a linux guy, but Windows is completely in my rear-view now.)
That is not acceptable in order to use a device you bought.
Also that means you're forced into a manditory arbitration clase and a class action waver.
https://en.wikipedia.org/wiki/Free_as_in_Freedom
-~`I do not want one ever`~-
But who would want that?
The stuff you have to agree to is a very significant cost. You're not just associating a username with a password.
I can get dogshit for free too. Doesn't mean I want it.