Topic: Driver for Windows on Arm

I have an AppleSilicon Mac installed.
Because I work with Windows, I installed Windows 10 and 11 using Parallels17.
I installed Driver for Windows but it does not work properly.

The installer shows "installation complete" but the driver is not actually installed and I cannot use the audio interface when I connect it.

I'm assuming that this is because the RME driver is not compatible with the Windows on Arm environment,
but what do you guys think?

Please help.

2

Re: Driver for Windows on Arm

What I can confirm is that we will never test nor support such a configuration. Sorry.

Regards
Matthias Carstens
RME

Re: Driver for Windows on Arm

I could understand.
Thank you.

4 (edited by idimata 2024-09-18 20:20:18)

Re: Driver for Windows on Arm

MC wrote:

What I can confirm is that we will never test nor support such a configuration. Sorry.

Now that the entire Windows ecosystem (Microsoft Surface, Acer, Asus, Samsung, HP, Lenovo, Dell) has gone Windows on Arm, will RME release native drivers for Windows' ARM-64 CPU architecture? Or at least will RME ensure compatibility with the emulation layer? I hope RME will at least consider it! I have three RME interfaces (UFX+, UCX II, and Babyface Pro FS), and I would like to use them on my new system. I hope the answer isn't that I need to sell it all and buy an inferior interface!

5

Re: Driver for Windows on Arm

You need to read some serious press to understand that the current generation of Snapdragon and Windows ARM computers completely fail (!) for Pro audio. Simply said - nothing works. You might ask again in 2 years...

Regards
Matthias Carstens
RME

Re: Driver for Windows on Arm

Not to mention the OP wants to use the interface in a Windows virtual machine on Apple hardware (now that is a perversion), which is not supported and never will be.

Fireface UCX II + ARC USB > ADI-2 Pro FS R BE > Neumann KH 750 DSP + MA 1 > KH 120 A

7 (edited by Kubrak 2024-09-19 10:18:55)

Re: Driver for Windows on Arm

Exactly MC! Latest WinOnARM trial, the third one, is as succesfull as the previous two. And x86 (at least by AMD) may deliver the same power efficiency as ARM without any compatibility issues on Win.... So, why anybody wants ARM on Win is beyond my comprehension capabilities.

Trials to push ARM to Win seems to be like trials to manufacture round square.

@idimata MS Surface 7 Pro (i5) runs with RME interfaces just great. It is passively cooled and one may get one secondhand for about 400 EUR. (Later models or i7 models are not passively cooled.) Even better is MS Surface 7+ Pro that has easily replacable SSD and i5 version is still passively cooled. Also it is last Surface that has classical CPU and not big.little....

8 (edited by idimata 2024-09-20 01:34:04)

Re: Driver for Windows on Arm

Kubrak wrote:

Exactly MC! Latest WinOnARM trial, the third one, is as succesfull as the previous two. And x86 (at least by AMD) may deliver the same power efficiency as ARM without any compatibility issues on Win.... So, why anybody wants ARM on Win is beyond my comprehension capabilities.

Trials to push ARM to Win seems to be like trials to manufacture round square.

@idimata MS Surface 7 Pro (i5) runs with RME interfaces just great. It is passively cooled and one may get one secondhand for about 400 EUR. (Later models or i7 models are not passively cooled.) Even better is MS Surface 7+ Pro that has easily replacable SSD and i5 version is still passively cooled. Also it is last Surface that has classical CPU and not big.little....

I already have a Surface Pro 7+ with the i5, with 5G, which I used for more than 4 years now. It performs very poorly and was the bane of my existence. It was holding me back. I couldn't even use certain programs I need to use. For example, DaVinci Resolve for editing videos for YouTube or otherwise: it's a no-bueno on my 7+. Certain plugins would struggle on it. Sampling libraries would be unstable and use huge amounts of CPU and Memory. Useless. Sure my RME works, but not much else.

The whole point of these devices is that they should be portable. If I have to carry an x86 Surface Pro to run my audio interface (and nothing else really) and an ARM-64 Surface Pro to run everything else, this is not going to work. I don't think that would work for most people, actually.

Edit:
Come to think of it, I actually also used an AMD laptop with a Ryzen 5, where I dual booted Windows and Linux (I did software engineering for >2 decades and have experience with both OS's). Linux is already ARM-compatible, and Windows's first foray into things ARM was terrible, but now the Windows experience is a lot better.

I would have to say that the difference between ARM-64 and x86 is night and day. Much faster, much more battery efficient, I don't hear any fans, and much more stable. I don't plan to go back. Even Linus Torvalds uses Linux on an ARM-based device. DaVinci Resolve 19 works on the Surface Pro 11, Reaper, and so many other programs I use.

I've done my research. It looks like to me like ARM-64 is here to stay, especially on Windows. I can easily see a lot of companies getting left behind and going the way of the dinosaur, unfortunately.

I think it may be best for RME to at the very least just ensure that it works via the Windows emulation layer for compatibility, even if there aren't any current plans to create native drivers. The Prism emulation layer appears to work very seamlessly for every program I've tried. There are just a few programs that don't work, really.

Re: Driver for Windows on Arm

OK, you may buy x86 MS Surface or other computer. MS Surface 7 Pro runs Traktor 3 Pro (DJ program) and it did run even few heavy CPU plugins. Of course not many. It is not computer for heavy work, it is Win tablet, not computer for heavy computing tasks.

And cannot you run RME interface in CC mode? I know no TM in that case...

Re: Driver for Windows on Arm

Kubrak wrote:

OK, you may buy x86 MS Surface or other computer. MS Surface 7 Pro runs Traktor 3 Pro (DJ program) and it did run even few heavy CPU plugins. Of course not many. It is not computer for heavy work, it is Win tablet, not computer for heavy computing tasks.

And cannot you run RME interface in CC mode? I know no TM in that case...

I do believe there should eventually be drivers and a compatible TotalMix experience. As a Linux user I can expect to be a second-class citizen and have to use class compliant mode, etcetera. But part of the reason why I still dual-boot Windows beyond compatibility is that I know that all the top companies are going to give it flagship attention. I don't expect to have the same issues on the Windows side. When Apple made the move to ARM, same thing. I think ARM is here to stay, and I don't think Intel's Lunar Lake and its battery life promises are going to steer it back to x86, because these CPU's will still have the same problems ARM solves including the thermal performance.

Anyway, this is just my opinion: All my years of building computers, programming them, using them... I've learned that my desktops and laptops are essentially the same components. I recently built a Mini PC with a 13th Gen Intel i7 for a project, and there was very little meaningful performance differences between the Intel Core i7-13700 used on desktops and the Intel i7-1360 I used in the Mini PC. They both have 13th gen CPU's. The memory also only differs slightly. There's just less room for expansion, less room for thermals, and less application for a powerful GPU. The essential ingredients are very similar. Of course we can into the technical nitty-gritty and argue about SoC's and the proximity needs for the memory to be close to them, etc! But my basic point is that the performance gains between tech of the same generation are going to be very roughly on par, except faster compared to the generation of tech before it.

In the same way, there's little performance difference in my experience between the Surface Pro line and a desktop I might build. The meaningful differences are going to be the form factor (it's portable!) and I/O, really. The Surface Pro 11 on ARM is faster than a 13th gen CPU. After getting the device, I can immediately tell the difference.

11 (edited by Kubrak 2024-09-20 16:35:47)

Re: Driver for Windows on Arm

But Intel does not have good times last few years and most probably will not have next few years.... So, comparing Intel x86 and ARM does not give right picture.... One has to compare AMD x86 and ARM if one wants to make fair comparison of those platforms.

Bigger computer/notebook may have better cooling so CPU may run faster clock... And RAM and SSD and GPU may run also better.

I guess Linux is still niche as well as ARM on Win, so no wonder developers do not hurry up to develop SW for them. Few customers,  50-100% more work and cost of support....  It does not make sense.  If WoA makes at least 20% marketshare of Win market, than maybe....

But I doubt WoA will succede in current, third trial. AMD CPUs may deliver the same without any compability issues. So why to bother with WoA, unless one likes to experiment with new gadgets.

Re: Driver for Windows on Arm

I hear what you're saying. However, I would say that it's new, I would not not quite classify it as niche yet. If it was just the Microsoft Surface line that went ARM-64 like with the Surface Pro lines that had the SQ1, SQ2, and SQ3 processors, I would say niche. However, this time all the major non-Apple manufacturers have gone ARM-64. So anyone who buys a reasonably decent non-Chrome-OS new laptop or 2-in-1 today will be on an ARM-64 device. I focused on Intel because they have  roughly 80% of the x86 CPU market, and AMD only roughly 20%. Of course I rocked an AMD CPU for quite some time, and they have been quite friendly to the Linux and FOSS community, but they too are being left behind. However, OEMs got on the Snapdragon X Elite bandwagon because there's a projection that 60% of their PC sales will be from Snapdragon X-based systems within 3 years -- that's a lot! They anticipate 30% of the PC market will be ARM-based within 2 years. Apple Silicon is already at 8-10%.

Windows 11 users are still only 31.6% of the marketshare compared to 64.14% on Windows 10. Of course it has something to do with the new higher hardware requirements for having Windows 11 that it's only 1/3rd of the current users, yet the pro audio market already has Windows 11 compatibility covered, even RME. (But it doesn't quite mean that Win 11 would be classed as niche, for example: Windows in total has 71% of the desktop marketshare and OS X only 15%.) If we were to compare all OS's including Desktops and phones, Microsoft is only second at 25% of all devices with the lions share going to Android at 45%, iOS third at 18.4% and OS X a distant fourth place at only 5.53%!!! However, major companies in pro audio still jump hoops to have MacOS compatibility, including RME. We're not going to be running audio interfaces from our phones, but I think you can appreciate the overall point.

I know it's a numbers game, but in 1-2 years, I don't think Windows ARM-64 will be only a small percentage of the users; I'm pretty sure you're going to see a lot larger of a number of individuals using Windows ARM-64 devices. I think it would become pretty awkward to not offer ARM-64 support. You could say that the previous Microsoft attempt at ARM was bad, but this time it's very different from the SQ# CPU days. Windows 12 is being worked on, and when it releases sometime next year it looks like that may do also do another market squeeze and push everything even more so towards ARM, although it may still have x86 compatibility.

13 (edited by Kubrak 2024-09-21 00:28:07)

Re: Driver for Windows on Arm

We will see, if ARM will succede to be widely used on Win. I strongly doubt it, but it is not uncommon that worse solution/technology wins above the better one.... So, maybe ARM will find the way to Win.

AMD x86 has the same energy efficiency like ARM, anyway it is RISC CPU inside and just has x86 frontend...

People are not on Win11, because they do not want them. I may install Win11 on five computers of mine, and MS tries to force me to install W11, but I do the best to avoid being forced to install them.... And I switched do Win10 a year ago, till then I was on Win7 which I like better than Win10.  I will switch to Win11 in two years or so. The later the better.

Re: Driver for Windows on Arm

Kubrak wrote:

We will see, if ARM will succede to be widely used on Win. I strongly doubt it, but it is not uncommon that worse solution/technology wins above the better one.... So, maybe ARM will find the way to Win.

AMD x86 has the same energy efficiency like ARM, anyway it is RISC CPU inside and just has x86 frontend...

People are not on Win11, because they do not want them. I may install Win11 on five computers of mine, and MS tries to force me to install W11, but I do the best to avoid being forced to install them.... And I switched do Win10 a year ago, till then I was on Win7 which I like better than Win10.  I will switch to Win11 in two years or so. The later the better.

I'm pretty sure ARM will overtake its competition, the Snapdragon X line also being RISC processors, and I don't think AMD's offerings will surpass the transition that's already taken place towards Qualcomm's chips, but I agree with you -- only time will tell! I agree with your actions/sentiment, I actually wouldn't be using Windows at all if Linux was compatible with everything for certain things I currently need to do. It's what we've got, however. I hope RME will strongly consider adapting support to ARM-64! We will see

Re: Driver for Windows on Arm

As MC said currently no, and they might reconsider it in 2 years, or so.

Pretty good backward compability is what is valuable on Win. And ARM breaks it. Win is not Mac where one company decides and nobody asks users or developers. It just comes and the rest of the "world" has to adapt to the new situation. On Win things are not as simple as that.

We will see if the third attempt to introduce ARM into Win will end up as two ones before, or it will depend better.

Re: Driver for Windows on Arm

Well I had the time to sit down and test things out. Of my 3 RME audio interfaces I tested my RME Babyface Pro FS in compliance mode, and Windows doesn't recognize the device: there are two warning symbols in the Device Manager. The audio interface isn't at all detected by the OS or any programs.

I understand what you're saying Kubrik, and I do get it having done software engineering in the past, especially developing software with device development on Windows and running into walls with many hours debugging. Regardless I don't think it's a great strategy to wait 2 years to see if things pan out and then start supporting people, in my opinion! I believe RME will lose business from sincere people (like myself -- I'm considering selling my audio interfaces, and it's shame because I think RME is the best out there!).

Re: Driver for Windows on Arm

If CC mode does not work on WoA, it is not RME's fault. It is fault of MS or designer of your computer...

If RME does not support Linux, which has higher marketshare than WoA, why should they invest their resources in WoA. And @MC clearly said that they have investigated it and found current WoA devices not capable work with audio.

Why should RME support platform that is not able to work reliably with their interfaces? It would bring no good. Just upset customers being angry that things do not work as they should.

Re: Driver for Windows on Arm

MC wrote:

You need to read some serious press to understand that the current generation of Snapdragon and Windows ARM computers completely fail (!) for Pro audio. Simply said - nothing works. You might ask again in 2 years...

https://devblogs.microsoft.com/windows- … n-windows/

It hasn't quite been two years but there was a lot of promising stuff announced at the Snapdragon Summit. At the very least there is reason to hope that Windows Arm machines will not completely fail for pro audio for much longer, and indeed are poised to get some features (such as in-box USB ASIO drivers) ahead of x86. Native DAWs are actually appearing.

Places such as Scan and DAWBench seem to be treating the news seriously.

Manufacturers such as Focusrite and Steinberg are preparing dedicated Windows Arm drivers and "other companies working on optimized Arm64 drivers for their audio products as well". Thoughts on these developments in general? Does it change RME's stance at all?

Re: Driver for Windows on Arm

idimata wrote:

Well I had the time to sit down and test things out. Of my 3 RME audio interfaces I tested my RME Babyface Pro FS in compliance mode, and Windows doesn't recognize the device: there are two warning symbols in the Device Manager. The audio interface isn't at all detected by the OS or any programs.

I understand what you're saying Kubrik, and I do get it having done software engineering in the past, especially developing software with device development on Windows and running into walls with many hours debugging. Regardless I don't think it's a great strategy to wait 2 years to see if things pan out and then start supporting people, in my opinion! I believe RME will lose business from sincere people (like myself -- I'm considering selling my audio interfaces, and it's shame because I think RME is the best out there!).

USB Audio Class 2 devices, if they are class-compliant, should work fine -- that's how class drivers work, and our existing USB Audio Class 2 in-box driver, albeit not really suitable for musicians, is the same code on x64 and Arm64. Our new class driver (the announcement this week) is expected to be identical, or nearly-so, across the two architectures as well. But neither driver has any real DSP or similar in it, which is where you start running into different libraries etc. across Arm/Intel.

Companies who want to create device-specific drivers and apps for Arm64 will be able to exactly like they do for Intel/AMD x64 today. We haven't changed anything to make that any different. I do recommend that any in-proc components be compiled as Arm64X fat binaries, however, to support both Arm64 and Arm64EC-compiled apps.

Other class-compliant audio interfaces I've tried work when plugged in to my Surface Laptop 7, even bus-powered ones, using our current in-box driver. Are you sure something else isn't going on, here?

Pete
Microsoft

20

Re: Driver for Windows on Arm

Psychlist1972 wrote:
idimata wrote:

Well I had the time to sit down and test things out. Of my 3 RME audio interfaces I tested my RME Babyface Pro FS in compliance mode, and Windows doesn't recognize the device: there are two warning symbols in the Device Manager. The audio interface isn't at all detected by the OS or any programs.

USB Audio Class 2 devices, if they are class-compliant, should work fine -- that's how class drivers work, and our existing USB Audio Class 2 in-box driver, albeit not really suitable for musicians, is the same code on x64 and Arm64.

Then Babyface Pro and UCX II should have worked when set to CC mode - they do under Windows 10 and 11. The in-.box driver has issues and limitations though, especially with higher number (than 2...) channel interfaces, so the UFX+ not working at all is the same on x86 and ARM platform.

@Idimata: unclear and misleading statements are not appreciated. In Pete's blog you clearly state that you tried BF Pro, UCX II and UFX+ in CC mode on a Surface Pro 11, which has Snapdragon CPU and runs ARM64. If none of them work there in CC mode then it is not us that need to work on that.

Regards
Matthias Carstens
RME

Re: Driver for Windows on Arm

MC wrote:
Psychlist1972 wrote:
idimata wrote:

Well I had the time to sit down and test things out. Of my 3 RME audio interfaces I tested my RME Babyface Pro FS in compliance mode, and Windows doesn't recognize the device: there are two warning symbols in the Device Manager. The audio interface isn't at all detected by the OS or any programs.

USB Audio Class 2 devices, if they are class-compliant, should work fine -- that's how class drivers work, and our existing USB Audio Class 2 in-box driver, albeit not really suitable for musicians, is the same code on x64 and Arm64.

Then Babyface Pro and UCX II should have worked when set to CC mode - they do under Windows 10 and 11. The in-.box driver has issues and limitations though, especially with higher number (than 2...) channel interfaces, so the UFX+ not working at all is the same on x86 and ARM platform.

@Idimata: unclear and misleading statements are not appreciated. In Pete's blog you clearly state that you tried BF Pro, UCX II and UFX+ in CC mode on a Surface Pro 11, which has Snapdragon CPU and runs ARM64. If none of them work there in CC mode then it is not us that need to work on that.

Please LMK if anyone has traces, crash logs, or otherwise finds that the RME interfaces, when in class-compliant mode, don't work on Windows 11 Arm64 devices. Like MC said, the current in-box driver has limitations (and is not suitable for music creation, really), but it should have been recognized and worked at a basic level.

Pete

22 (edited by ramses 2024-10-28 10:28:29)

Re: Driver for Windows on Arm

idimata wrote:

Windows 11 users are still only 31.6% of the marketshare compared to 64.14% on Windows 10. Of course it has something to do with the new higher hardware requirements for having Windows 11 that it's only 1/3rd of the current users, yet the pro audio market already has Windows 11 compatibility covered, even RME.

It's no wonder that Windows 11 is so badly received.

Keyword hardware requirements: Virtualization-based security (VBS) is so poorly designed and costs so much performance that CPUs manufactured before 2017 can no longer be used. In any case, there have also been massive complaints from the gaming community, which, like the recording industry, needs optimal performance without the VBS slowdown.
It is completely unacceptable that systems are slowed down so much by poor design that hardware replacement is unavoidable, and even the newer systems do not achieve the performance they could deliver.

Basically, this is an environmental scandal beyond compare. Why do we have to dispose of so many systems worldwide for a safety feature that certainly does not create absolute safety, which will never exist?

Specifically, my situation: why should I throw my two Xeon E5-1650v4 and E5-1680v4 based systems for Windows 11 in the bin? The CPUs are so powerful that they can play 400 tracks with 800 VST without cracking and also play games like Cyberpunk 2077 Phantom Library, Black Myth Wukong or X4 foundation in Full HD using RTX4070.

Furthermore, Windows 11 is a "permanent construction site". Too many things are changing. Among other things, Microsoft has also removed things that later had to be reinstalled after complaints from users, including too much back and forth in this area. I wonder why proven concepts were removed or changed at all.

Especially the new administration menus are also just a nuisance. Every so often you have to scroll back and forth so much because the menus are simply badly designed. Some fonts are too large, and the spacing is too high, or too much has been packed into one menu. My guess: that's what you get when you give developers 4k monitors. Nobody thinks about the people who still use full HD anymore, so that the hardware requirements for games don't increase immeasurably.

Finally. I also don't trust Windows 11 systems according to privacy. With every release of Windows, EULAs have been modified, in which Microsoft grants itself more and more rights.

Especially, the implementation of AI is particularly daunting. Windows 11 uses AI to continuously analyze screenshots, storing content locally for up to 3 months—ideally on an AI chip, otherwise on the CPU. Data like text inputs and usage patterns can be uploaded to the cloud, where it may be analyzed for personalized ads. According to the EULA, anonymous data may be shared with advertising partners, raising privacy concerns for many users.

Microsoft claims strong encryption and local storage for Windows 11 AI data, with optional cloud use and user control. However, it's unclear how consistently data remains local and whether anonymous data may be indirectly used for advertising purposes.

In short: all in all, I see nothing but disadvantages in terms of performance and data protection by switching to Windows 11. I tried Win11 for a while on a 2nd disk but removed it for the above-mentioned reasons.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub14

Re: Driver for Windows on Arm

Exactly Ramses.

I may install Win11 on almost any of my computers (10 or like), but why should I go for something worse than I have now? I will stay on Win10 as long as possible....

I would not "upgrade" to Win11 now, even if Microsoft pays me for that move.

Re: Driver for Windows on Arm

ramses wrote:
idimata wrote:

Windows 11 users are still only 31.6% of the marketshare compared to 64.14% on Windows 10. Of course it has something to do with the new higher hardware requirements for having Windows 11 that it's only 1/3rd of the current users, yet the pro audio market already has Windows 11 compatibility covered, even RME.

It's no wonder that Windows 11 is so badly received.

Keyword hardware requirements: Virtualization-based security (VBS) is so poorly designed and costs so much performance that CPUs manufactured before 2017 can no longer be used. In any case, there have also been massive complaints from the gaming community, which, like the recording industry, needs optimal performance without the VBS slowdown.
It is completely unacceptable that systems are slowed down so much by poor design that hardware replacement is unavoidable, and even the newer systems do not achieve the performance they could deliver.

VBS requires specific support in the CPU. This is like saying DirectX is poorly designed because without a discrete GPU, it's slower. VBS is another security measure to help protect your computer, and it's also in Windows 10. And, if you don't want to use it, you can turn it off. But most people who think they don't need it, should probably be using it. It's one of the only ways to protect against kernel-model malware.

For years, virus and malware authors had nearly free rein on Windows. It resulted in a bad reputation and annoying third-party subscription AV products that would tank PC performance. So Microsoft started putting into the OS a lot of security measures like this and they have helped. Meanwhile, infection rates in other operating systems and phones are going up and they're having to implement functionally similar approaches.

Especially, the implementation of AI is particularly daunting. Windows 11 uses AI to continuously analyze screenshots, storing content locally for up to 3 months—ideally on an AI chip, otherwise on the CPU. Data like text inputs and usage patterns can be uploaded to the cloud, where it may be analyzed for personalized ads. According to the EULA, anonymous data may be shared with advertising partners, raising privacy concerns for many users.

No. It doesn't. The screenshot bit is Recall, which is available only on Copilot+ PCs, and is off by default. It's not uploading to the cloud for analysis or used for ads or anything like that. It's so you can make a query like "What was that blue car I saw in a photo a month ago?" But, again, it's opt-in. The recent news stories about this don't understand what a provisioning/feature package is in Windows and how it may or may not relate to something actually running on the PC.

Anyway, this is a forum for RME products, and this thread is specifically about the Arm driver, not about opinions on an operating system.

Pete

25 (edited by ramses 2024-10-28 20:51:38)

Re: Driver for Windows on Arm

I apologize if you found my comment off-topic, but I was responding to a user's post, and I believe it does have a relevant connection to the discussion.

I think Microsoft should genuinely care about why customers are hesitant to migrate to Windows 11. After all, the success of ARM on Windows is tied to this issue, especially since it's unlikely that any improvements for ARM will be backported to Windows 10.

While I understand the importance of security, it shouldn’t become a catch-all excuse for implementing changes that ultimately lack acceptance.

Moreover, in order to disable VBS, one would first need to be able to install Windows 11, but that’s already being blocked by the installer (for hardware which is only 7+ year old, but still performant).

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub14