I sure hope it will come with SystemReady or a similar standard for something UEFI-like instead of the custom per board image many ARM devices need currently.
It would be an instant buy for me, especially since it wouldn’t be as locked down and work better with Linux. I’ll deal with the software issues later, gimmie now!
Hopefully the open source Qualcomm drivers will support this chip. The SDM845 chip is pretty well supported on mainline kernel, using SDM845 with 6.5 kernel and postmarketOS to type this comment. The Freedreno driver is probably the best ARM GPU driver in Mesa.
Exactly. Thanks to raspberry pis and other ARM SBCs there’s been a lot of ARM native support on Linux. Windows really hurt themselves with their initial ARM support.
Windows already works on ARM, but some X86 programs are still a little slow. The Snapdragons previously used weren't as fast as the Apple silicon too, so that didn't help. The original Surface X did have a lot more problems, but Surface 9 on ARM from all reports works well.
From a security perspective, as long as you check the hash against Microsoft’s website then it should be okay. Otherwise I’m not sure where to get Windows on ARM ISO’s from.
Both are instructions sets. They are part of the equation that gives a CPU its “power”, but it isn’t the only reason.
What gives ARM its power efficiency edge is its smaller instructions set, which translate to smaller die size to do the same work, which is also its Achilles heel as it means that some workload that uses those missing instructions need to be either translated by the hardware or the software, or it will just not work. Both have their own inconvenience (bigger die size and less energy efficiency or bigger overhead and slower execution.
But for workloads that do not use x86’s specificities, ARM is very competitive.
No, and the above commentor is a little mixed up. While we originally thought the benefit of RISC CPUs was their smaller instruction set - hence the name - it’s turned out that the gains really come from a couple other things common to RISC architectures. In x86 pretty much every instruction can reference memory directly, but in RISC architectures you can only do it from a few specific instructions. Modern RISC architectures actually tend to have a lot of instructions, so RISC means something more like “load/store architecture” nowadays.
Another big part of RISC architectures is they try to make instruction fetch+decode as easy as possible. x86 instructions are a nightmare to decode and that adds a lot of complexity and somewhat limits optimization opportunities. There’s some more to it, like how RISC thinks about the job of the compiler, but in my experience load/store and ease of fetch+decode are the main differentiators for RISC.
More towards your question, a lot of the issues with running x86 programs on ARM (really running any program on a different architecture than it was compiled for) is that it will likely depend on very specific behaviors that may not be the same across architectures and may be computationally expensive to emulate. For some great write-ups about that kind of thing check out the Dolphin (Wii emulator) blog posts.
Whoever can make a compatibility layer that successfully translates x86/64 to arm and vice versa and make it widely available will be a major player in the market. Valve has already somewhat done something similar with proton and Apple with Rosetta 2.
I guess it depends on how you define “real processing power”. I run Windows on Arm on my Mac Studio through Parallels. I installed Steam and played Civ 4 and it’s great. Sure, it’s an old game but it runs smoothly.
It’s gotta be Microsoft building it into Windows. The Apple Silicon transition wouldn’t have been nearly as smooth if people had to pay for, say, CrossOver or something to use their Intel apps. And the tepid response to the ARM Surface models makes me think that it’s a must, despite the UWP dream.
The good news is that Rosetta 2 shows it can be done extremely well!
Apple developed it as a stopgap. In the Windows world x86/64 will be around for a long long time. Not sure if anyone is willing to support something like that for the next 10 years.
It’s all a question of market share. If (big if) arm gets a foothold into the Windows market, software vendors will simply offer two binaries and/or Microsoft could offer tooling to offer easy porting.
Apple’s real genius move though is not Rosetta, but including x86 compatibility features into the Mx chips. That way the emulation is much faster.
Windows does have a 32/64 bit x86 compatibility layer and most of what I’ve seen through limited bashing around in VS2022 leads me to believe that it has arm as a fairly targetable build target already.
I could be wrong but that might be Linux only. Windows and MacOS both have their own proprietary compatibility layers, but Windows had shit support for theirs for years which hurt their reputation badly.
Proton and Rosetta 2 are two totally different beasts. One allows windows programs to run on non-windows hosts and one translates x86 to Arm.
I’m not aware of Proton doing anything like Rosetta 2 and if it did Steam would have probably used an Arm chip in their Steam Deck instead of an x86.
Maintaining 2-way compatibility doesn’t seem like an important goal. One way, x86->Arm, sure but not Arm->x86. Apple clearly sees x86 as a dead end for its own product lines and we will see if the rest of the industry follows suit over time. Of course there is a ton tied up in x86 but aside from legacy apps or games I don’t have much need of x86 in my life.
Even the servers I run are trending towards Arm due to the power savings. AWS graviton stuff is like ~25-30% cheaper than x86 last I looked
I understand proton isn’t the same thing, it was just an example of a compatibility layer…and how would a bidirectional compatibility layer not be beneficial? X86 in servers may be going away and even that’s debatable, but x86 isn’t going anywhere in the consumer space. Graviton chips are great, but they’re useless if there’s no viable way to translate those x86 legacy applications over to ARM without breaking the bank until your business is ready to transition the workload to ARM.
Amazon was working on a compatibility layer specifically for this purpose, however I suspect they’ve given up because they’ve slowly added Intel and AMD chipsets back into their general purpose ec2 class for newer generations and there hasn’t been a single word about compatibility with graviton other than just use arm based workloads.
You just can’t move to ARM because it’s cheaper, that’s just not going to work. You need to make the effort to move away from x86 and adopt applications that are arm native before making that jump. With a compatibility layer it doesn’t matter, that’s where the money is, if I can build a compatibility layer that translates an x86 binary to an arm binary, then I can move those workloads to the cheaper and more efficient server class.
It’s a different problem under the same name. You’re comparing apples to oranges by comparing WINE and Rosetta.
WINE (as I understand it) explicitly does its best to maintain the instructions and basically just implements functions that are missing.
Rosetta actually changes out every instruction (and presumably has to do a WINE-like thing as well to handle dynamically linked code, though in theory this is more of a 1-to-1 mapping since you’re not swapping OSes).
if there’s no viable way to translate those x86 legacy applications over to ARM
For a properly written application that doesn’t have a lot of third party binary dependents… This really shouldn’t be that big of a job. You just recompile the code on ARM, test, and fix any bugs.
I’d speculate that most of the time this stuff comes up like: “this was written for Windows with some very specific Microsoft libraries/tech in mind and we want to run it on Linux and ARM as cost cutting measures.”
Granted, it’s not unusual for a large enough code base to contain some code that wasn’t written in an architecture neutral way/does some shady stuff.
I’m not comparing Rosetta and proton…I’m just using them as examples of compatibility layers that make their products viable. I frankly don’t give a shit about the inner workings of Rosetta or proton, but thanks for clarifying I guess.
Add comment