“Windows 11 26H1” is a special version of Windows exclusively for new Arm PCs
- Thread starter JournalBot
- Start date
Upvote
51
(53
/
-2)
Sounds like it. From Microsoft's annoucment:
https://techcommunity.microsoft.com...to-know-about-windows-11-version-26h1/4491941
It sounds like they want to give themselves a year to address any bugs that their changes to accommodate the Snapdragon X2 Elite chips might have on other hardware. Or hell, there may even be bugs on the Snapdragon X2 Elite hardware, but the impact will be limited to just that hardware, until they merge it back into the mainstream release.
This reminds me a bit of when Android 3.x was a tablet-only OS, and it wasn't until Android 4.x that the tablet and phone lines merged again.
Upvote
54
(54
/
0)
I'm going to be the whiny and annoying one and say it outright, my interest in Windows and Microsoft is and will remain dismally low for as long as they don't fix the disaster of bugginess, forced updates, and AI integration that Windows 11 has become. I don't know where things went wrong specifically (I mean I do, but that's beyond the scope of this comment) but something has to change. Using Windows 11 is an outright unpleasant experience with how hostile to the user it is. I have never really hated a Windows version before - I was one of the handful of Vista stans back in the day, and I even managed to be excited for Windows 8 in theory if nothing else. But Windows 11 has passed the rubicon. Microsoft needs to strip out the bloat, stop vibe coding a fucking OS, get a better handle on QC, stop forcing AI, and most importantly, respect the user and their preferences. Then, and only then, am I interested in picking up a Windows device again.
Holy shit
Holy shit
Upvote
26
(41
/
-15)
I read elsewhere that 26H1 is just 25H2 with very specific changes for the new ARM processors. Considering even my 5 ARM PC users won't be getting this release, I'm not giving it a second thought until the 2027 release which is when we're supposed to be back to a single release.
Upvote
6
(7
/
-1)
We've got a few dozen devices with Snapdragon chips now and I've been pleased with Windows on ARM so far. End users love the battery performance. The main issues have been a lack of driver support for some peripherals, printers and scanners mostly, and with running some memory-intensive legacy apps under emulation in Prism. Not entirely sure how the emulator should respond to something needing more address space than should be available to an x86 app, but it seems like "let's just die and try again" is pretty typical.
Upvote
29
(29
/
0)
Upvote
36
(37
/
-1)
I run Windows 11 on four devices and haven't dealt with any bugs. None of them are ARM devices though, so is that where you've had trouble?
Generally, the AI stuff seems to me to be more of a wasted dev effort than something that actually affects me. The operating system is stable, performant, and pretty user friendly (I've not had to use control panel in five years!).
It's not perfect, but it seems like a pretty similar experience to my MacBook in the areas I most care about for an operating system.
Upvote
0
(17
/
-17)
I mean there would be none if Windows wasn't a -censored-.
Been daily driving Linux Mint for 6 months now, quite happy with it, 2 days ago I had to boot up a virtual PC to troubleshoot an app issue - opened the start menu, typed in 4 letters of the app name (the ONLY non-built-in app currently installed) then watched how it took 4 MINUTES to show it as a result...yes, I actually timed it since I powered it off and tried it again from the original boot state...and stuff like this has been happening to me for years outside of virtual machines too.
Why would anyone tolerate this unless absolutely forced to by their job?
Upvote
3
(13
/
-10)
I have a Surface laptops with a Snapdragon X Elite processors in it, and it's definitely developed some quirks these past few months that I've never seen in any other Windows device. Most annoyingly is sometimes it'll get stuck in a smart charging mode where it never actually charges, regardless of whether or not I'm using the provided charger. Another frequent one is the onboard speakers will be at about half their usual volume, so 100% volume sounds like 50%, 50% sounds like 25%, etc. A reboot resolves these issues, but it's still annoying. None of these things happened in the first year or so I owned it, so I assume there were issues caused by the 25H2 update, necessitating this out of cadence update.
Upvote
11
(11
/
0)
What's interesting is that your device isn't eligible for this update. The Microsoft announcement specifies that it will only be available on new hardware with " select new silicon." And this article says it will be on laptops with Qualcomm’s recently announced Snapdragon X2 Elite chips.
Upvote
15
(15
/
0)
Oh, my apologies, that's a total brain fart from me, I somehow completely misread that from the article. I guess this means that there's something unique to these newest generation that came up late in development that Microsoft needs to address. That does make me kind of bummed that the issues I'm seeing might not get fixed, then.
Upvote
7
(7
/
0)
It wouldn't be a new version of any software if it didn't. Changes mean risk of bugs. It's not a Windows (or Microsoft) specific "feature".
Hell, I just found a neat little bug in a new feature I just added to my own crappy piece of software (and along the way while reviewing things, found another completely unrelated one that had been sitting there for a while, just never triggered.)
Upvote
5
(6
/
-1)
The 3.2 kernel is quite old, though, initially releasing in January of 2012
Upvote
23
(24
/
-1)
I've noticed some weird issues with USB-C PD charging on my absolutely bog standard x86 based Windows 11 laptop, so I doubt it's related to being ARM or anything. For no reason at all I can see (it's the same charger it always uses, the vendor provided one), it decides to sit in a lower charging mode so instead of actually charging the battery it just slooooooowly discharges. And then a sometime later it decides "nope, the same charger is grand for fast charging". A bit frustrating to say the least.
Upvote
5
(5
/
0)
Mikrotik rOS 6! (yes, it uses the 3.2 kernel. It's on "we'll support it as long as we have to so pleeeeeease migrate to 7" long term support). Unfortunately for me, 7 requires x86-64 hardware so the poor little ALIX PC is just sitting there on rOS 6.
Edit: Seeing this is actually a Windows article and Windows 11 dropping support for CPUs is a big deal ... I'm not actually complaining about another vendor doing something similar - Mikrotik dropping support for 32 bit x86. It's an observation, not a whinge
I think it's fair enough in this case, and I've made that clear to Mikrotik support when I've filed bugs that "this is just a heads up in case someone is bored. Don't care too much about it because I don't"
Last edited:
Upvote
3
(3
/
0)
My guess is that it’s related to an optimised version of Prism for the new X2 processors, which is the emulator that allows x86-64 code to run on ARM processors.
Upvote
1
(1
/
0)
Thankfully there's no A/V workstations for sale with ARM because I don't want to have to test on yet ANOTHER Windows version.
Upvote
2
(2
/
0)
Nvidia N1X Processors are also going to receive this supposedly.
Anyways what's actually going on:
Windows 25H2 and indeed 26H2 (when its released) are actually still using the 24H2 platform. 26H1 is a new platform and should merge in with 27H2. I'm sure the reason they aren't just coming out and saying that is that they might opt in to calling 27H2 Windows 12.
Anyways what's actually going on:
Windows 25H2 and indeed 26H2 (when its released) are actually still using the 24H2 platform. 26H1 is a new platform and should merge in with 27H2. I'm sure the reason they aren't just coming out and saying that is that they might opt in to calling 27H2 Windows 12.
Upvote
4
(4
/
0)
Same here! I thought Windows Vista was almost as good as sliced bread. Sure, Windows 7 was better, and I was extremely excited for it, but I was excited precisely because it was Windows Vista, just better.
I've not had much experience with Windows 11 because I left the Windows camp before its release, but my few interactions with it have indeed been painful almost to the point of rage-inducing. Combine that with forced obsolescence and Microsoft are really not making themselves many fans lately.
Edit: typos
Last edited:
Upvote
7
(10
/
-3)
Does this announcement change any Windows 12 speculation?
The new schedule for ARM editions make me wonder if they wanted a Windows 12 this year but decided to delay it. PC component price hikes due to datacenters would give a 2026 Windows release sluggish sales.
The new schedule for ARM editions make me wonder if they wanted a Windows 12 this year but decided to delay it. PC component price hikes due to datacenters would give a 2026 Windows release sluggish sales.
Upvote
1
(2
/
-1)
Interestingly, two of the three co-founders of NUVIA resigned from Qualcomm this month.
Gerard Williams III and John Bruno both resigned from Qualcomm last Monday. Microarchitectures take 2-3 years from design to shipping, so they likely contributed to future Qualcomm SoCs, but it is telling they are leaving Qualcomm before Windows on Arm gained any major momentum.
Williams was the architect of the M1 cores. I'm sure he and Bruno will find gainful employment quickly.
Upvote
9
(9
/
0)
What do Intel / AMD do with new CPU generations, that seemingly Qualcomm cannot replicate? See Panther Lake.
This is not the first rodeo for Windows on Arm. This is the 9th year for Qualcomm's Windows on Arm SoCs (the good old Snapdragon 835 Mobile PC Platform) and the 14th year for Microsoft's Windows on Arm release (Windows RT).
I would've thought Arm was a first-party ISA for the OS.
Upvote
1
(1
/
0)
Upvote
9
(9
/
0)
Microsoft stopped reporting Surface revenue individually a couple of years ago after multiple heavy losses in 2023. Surface was lumped into Devices. Then in 2025, they stopped reporting Devices individually, too, now lumped into Windows OEM.
What I can find: Sometimes they report %, sometimes a hard number.
2021: Surface baseline: $6.7 billion in revenue
2022: Surface is up 3% → $6.9 billion in revenue.
2023: Devices is down $1.8 billion → $5.1 billion in revenue
2024: Devices is down $815 million → $4.3 billion in revenue
2025: Devices are down again, no longer reported
Microsoft has money, no doubt, so Surface is hopefully safe. But I am unsure if any Windows OEMs even use it as a reference device. It seems more and more OEMs prefer to model MacBooks rather than Surface (which itself was modeled after MacBooks).
What I can find: Sometimes they report %, sometimes a hard number.
2021: Surface baseline: $6.7 billion in revenue
2022: Surface is up 3% → $6.9 billion in revenue.
2023: Devices is down $1.8 billion → $5.1 billion in revenue
2024: Devices is down $815 million → $4.3 billion in revenue
2025: Devices are down again, no longer reported
Microsoft has money, no doubt, so Surface is hopefully safe. But I am unsure if any Windows OEMs even use it as a reference device. It seems more and more OEMs prefer to model MacBooks rather than Surface (which itself was modeled after MacBooks).
Upvote
4
(4
/
0)
That lines up very well. I wonder how many of the ~200-person NUVIA team also cashed out as soon as financially prudent. Manu Gulati, the 3rd co-founder, has stayed.
Looking at their timeline, NUVIA has been at it for 12 years. That is a long time on a single project.
//
I hope Qualcomm keeps at it. Apple, Intel, and AMD all will need a strong CPU competitor for the next few decades and Qualcomm is the best competitor so far (I guess we'll see what NVIDIA does, if ever).
Upvote
4
(4
/
0)
Yeah, done that. First thing I've done (it's amazing how loose a really old USB C connector on a PC can get, my 2015 MacBook still charging is a miracle, but its usually the cable thats first to go). I think they've tried to update the state machine for USB-PD versions and typoed (or swapped something) somewhere so every now and again it misfires. It'll be fixed sooner or later, I'm sure.
Upvote
3
(3
/
0)
Upvote
-9
(5
/
-14)
Windows has historically supported many architectures. Of course for many years both x86 and x86_64 were supported, but also MIPS, Alpha, PowerPC, and 32-bit ARM (and probably others I forget) were supported at various times.
Supporting lots of architectures has real costs, and for a commercial entity like Microsoft if the money isn't there it makes sense to drop it.
Upvote
17
(17
/
0)
It should run in UTM on my Mac Studio M2 Ultra, all the previous ARM Windows have run very well.
Upvote
1
(1
/
0)
WinCE had a whole collection too, the most amusing being the SH3 variant that was ported to the Sega Dreamcast. Games themselves that booted directly off the disk didn't actually have to run under Windows CE of course, and pretty much didn't. But there was Windows of sorts for it.
Upvote
7
(7
/
0)
Surprised people don’t remember this! Windows NT supported multiple architectures.
Also, Microsoft released a special version of Windows XP for x86-64 PCs.
Upvote
7
(7
/
0)
My IT shop has largely avoided ARM for anything beyond testing and for in house devs. Adding a whole separate version of windows that will require testing does not bring any confidence into me either. Guess we will continue to be in "wait and see" for another year on these.
Will say the ARM based macs have been solid so far though. Do run into the occasional hiccup here and there, but the user base for those are largely in house editors and the occasional exec that wants one.
Will say the ARM based macs have been solid so far though. Do run into the occasional hiccup here and there, but the user base for those are largely in house editors and the occasional exec that wants one.
Upvote
1
(1
/
0)
After years of half-hearted attempts and relative neglect for ARM, this would normally feel like welcome news. If only we felt like the added attention would be positive in the current tech climate.
Upvote
-1
(0
/
-1)