ARM64

I've been using NCrunch for many(!) years now. It looks like the next laptop I'll be using will be an ARM64 processor on Windows (surface laptop 7). It appears ARM64 isn't currently supported by NCrunch. Is this on the roadmap? If not, can I get it added? Even if this worked in emulation I could get by, but posts so far suggest it doesn't work at all.

Thanks
Jason

Hi Jason,

Thanks for posting. As much as I would like to see NCrunch working on ARM, realistically I don't think this is something that's going to happen. We've overextended ourselves somewhat with the number of toolsets and platforms we already support, and adding a new one would be outside of what we could justify maintaining even if we could find the budget to implement it.

As much as I understand the circumstances that might lead to that desicion it's also sad to read that "fact"(?) on the other hand.

Since I never give up, I just want to understand some things:
Where would you see the most of the amount while adding support for ARM64?
Is it the development part it self?
Is it the testing part after having developed the architecture support before being able to release it to public?
Is it the support after having it released?

Maybe you could introduce some "beta program" where people could test prereleases and find bugs earlier.
Also I just saw you're still supporting VS down to version 2010 (do people still even use it??). Could cutting support for older versions be a solution to free ressources maybe?

Edit: btw: is it possible to change my account name in this forum?

Is it the development part it self?
Is it the testing part after having developed the architecture support before being able to release it to public?
Is it the support after having it released?

All of the above.

It's impossible to estimate the effort required for something like this, because ultimately it involves fixing everything until it works. We simply don't know how many hangups will exist, and it could be months into the work before realising that it was a terrible idea to try.

We are integrated at so many levels of the platform that regardless of how well the platform handles adaptation to ARM, we will have stuff that slips between the cracks.

It also means that assuming we do implement it, the entirety of the product needs to maintained on it.

Unfortunately there is simply no where near enough demand for this to justify the time and effort required. Basically, I'd be setting myself up for a potentially tragic loss that could even sink the product entirely. It would be rather irresponsible, actually.

Also I just saw you're still supporting VS down to version 2010 (do people still even use it??). Could cutting support for older versions be a solution to free ressources maybe?

Removing support for the older versions of VS will actually cost more time that it will save at this point. We would need to migrate a lot of tests and environments to do this. With VS2008, it was worth it because there was no WPF UI for it, so we couldn't upgrade parts of the UI without considering it. VS2010 has the same basic UI as VS2022, though some areas of the integration and core platform are clearly very different.

Edit: btw: is it possible to change my account name in this forum?

Unfortunately not, but it IS possible to set up a new account and transfer all details over to it (such as email, licenses, etc).

Thanks Remco. I know support for this really comes down to a guess about the market penetration of ARM on Windows. It's hard to predict that, although some crazy things are afoot. All Microsoft's newest Surface laptops are Qualcomm. Unless they backtrack on that it seems like ARM on Windows could break out of being a niche product. That was a pretty bold move by MS.

Remco wrote:

Is it the development part it self?
Is it the testing part after having developed the architecture support before being able to release it to public?
Is it the support after having it released?

All of the above.

It's impossible to estimate the effort required for something like this, because ultimately it involves fixing everything until it works. We simply don't know how many hangups will exist, and it could be months into the work before realising that it was a terrible idea to try.

We are integrated at so many levels of the platform that regardless of how well the platform handles adaptation to ARM, we will have stuff that slips between the cracks.

It also means that assuming we do implement it, the entirety of the product needs to maintained on it.

Unfortunately there is simply no where near enough demand for this to justify the time and effort required. Basically, I'd be setting myself up for a potentially tragic loss that could even sink the product entirely. It would be rather irresponsible, actually.

Also I just saw you're still supporting VS down to version 2010 (do people still even use it??). Could cutting support for older versions be a solution to free ressources maybe?

Removing support for the older versions of VS will actually cost more time that it will save at this point. We would need to migrate a lot of tests and environments to do this. With VS2008, it was worth it because there was no WPF UI for it, so we couldn't upgrade parts of the UI without considering it. VS2010 has the same basic UI as VS2022, though some areas of the integration and core platform are clearly very different.

Edit: btw: is it possible to change my account name in this forum?

Unfortunately not, but it IS possible to set up a new account and transfer all details over to it (such as email, licenses, etc).

That's sad to read an I wish u all the best!

Just to add weight to this request, my next work laptop was also going to be an ARM64-based Surface laptop. Every piece of software I checked either worked natively, or would be supported by emulation, except NCrunch. I love NCrunch and would struggle to work efficiently without it, so have requested an Intel/AMD laptop. But I think the tide is turning on ARM64. Pressure will build. 😅

I'd like to add my weight to the request too. I've been a Mac user for nearly as long as I've been an NCrunch user, and initially I thought I'd be unable to use VS again with the release of the Apple Silicon-based models. It seems though that getting a Visual Studio setup running on a Parallels VM is ever more feasible with the availability of VS on ARM64—I was pleased to see ReSharper supports this environment. To be frank, I don't enjoy using Rider the way I do VS+ReSharper+NCrunch, and JetBrains' continuous testing feature is way off the experience NCrunch provides. I'd be happy to offer feedback on any private builds you might have produced :)

Windows on ARM64 recent adopter here (surface 7 laptop, x-elite, 64GB RAM). Switch between VS2026 insiders and Rider. On the whole been a smooth experience I'd have to say with a couple outliers (e.g. SQL Server container is x64 only). Missing NCrunch obviously.