Actually a lot of companies rely on emulation -- if it allows them to keep running the software they want to run. I've had to set up things for multiple companies over my work lifetime. The biggest reason is that critical software costed WAY more than an emulation platform and would cost WAY more to get it rewritten, and they'll always take the cheaper option. That said, Apple isn't known for thinking backwards compatibility is important, so I wouldn't rely on rosetta..
Is M1 base really that superior to Intel i9 or i7?
- Thread starter TheOutlier
- Start date
- Sort by reaction score
If you want a Maserati, just buy one.
mi7chy
Suspended
You must be thinking of virtualization and not emulation.
The problem with Rosetta 2 is that it will eventually go away. This is very difficult to plan for for customers if they pull it in a release and you need to support that release or your employees need to use it.
throAU
macrumors G4
If you're relying on software that is unmaintained by the vendor (to the point where they haven't even bothered to port to the current architecture in multiple years via little more than a re-compile) you potentially have more serious problems.
Whether you're using an intel Mac, PPC, ARM/Apple Silicon or a Windows or Linux machine.
It's not just Rosetta being pulled than can cause this. Library deprecation, either in MacOS, Windows or Linux can and does break stuff. Just ask all the devs (including within Microsoft) who were relying in Silverlight in their apps.
The brokerage world doesn't care about macOS. It's an afterthought.
They update the software. The just don't care about macOS. It's run through WINE and then Rosetta. So it's possible to run it on macOS. Right now. I don't expect them to do anymore given that the haven't done anything more than make sure that it runs via WINE. If you wanted to run it in the PPC days, you ran Windows under emulation. A truly awful experience.
Fidelity has a lot of captive customers. They do 401ks for a ton of large corporations. They do pension and 403bs for lots of universities, hospitals, schools and non-profits. You, as an employee, don't get a choice in vendor. Not that it would matter that much. Areas of financial services is just a mostly Windows world.
Madhatter32
macrumors 68000
The problem with Fidelity is that they claim that their proprietary trading platform is Mac compatible when it barely runs. It's an embarrassment and make them look like an amateur outfit. TDA is way ahead of them in terms technology. What Fidelity needs to do is buy a third party platform -- say Lightspeed -- and enter the 21st centrury. This is exactly what TDA did with ToS. It just goes to show that an i9/i7 based Mac is most definitely a better option than an M1 Mac for a serious trader. That said, the OP's stated software needs suggest that an M1 would be the better device for their use case.
TDA has a big performance problem running on Apple Silicon but there's a workaround. But it's somewhat of a pain to do. You download the Linux version and do some hacking of the installion kit jar files and run it directly from Java. Lots of M1 users have talked to TDA's customer service asking them to fix it so that they can just run it well without Rosetta 2 and the resulting performance hit.
This is why I have four computers on my desktop including an M1 mini. It allows me to run just about everything.
januarydrive7
macrumors 6502a
You're confusing translation and emulation. Rosetta 2 performs a one-time translation from x86 to arm, unless there are specific dylibs that cannot be translated statically. In those cases, it uses a just-in-time translation when dynamically linking. It's not only a lot more sophisticated than Rosetta v1, but more efficient, and completely built in house (not licensed), so there's not reason to think it will necessarily go away in the same time-frame Rosetta v1 disappeared, as there are no license fees to keep it around indefinitely.
The expectation of it going away is a business decision to force software companies to do native ports.
There are companies that rely on emulation. If your original computer was really slow, then emulation may be a viable way to run software for an architecture where the hardware is no longer made.
throAU
macrumors G4
I'm sure the feeling is mutual.
Get a Windows box for Windows jobs.
Fitting square pegs into round holes is never going to work out well for you.
Madhatter32
macrumors 68000
You know, I keep hearing people make that argument. The fact of the matter is that Apple is in the computer business. Computers are not like musical instruments -- with each instrument playing its part. They are intended to perform a whole host of operations for many different use cases. No company should view its computer business as servicing one particular use case -- say video editing. That would be a pretty dumb business plan. Instead, they should, and usually do, focus on the broadest possible and best user experience.
Indeed, but emulation support can (and almost certainly will) be withdrawn at some point, so relying on it is a "last resort" to keep the lights on. It will buy you some time while you hope that either the original software vendor has upgrade path, or if not, as is often the case, for you to find a replacement for the software.
I've worked a several migration projects where a customer's legacy software is no longer supported, which can be a huge risk for the company when they realise there is no vendor support, and that their own internal support will gradually erode due to natural employee attrition.
I don't know - what about Avid / ProTools and quite a few other specialised AV products that can really only do one thing? They're not mass market products, and it's not exactly a "computer business", but there is a case for specialised hardware.
Even in mainstream IT, there is clear division between the kinds of software you run on a client machine (desktop/laptop/workstation), a Linux server, a Windows Server and other platforms such as IBM Power, SPARC etc. Plus specialised GPGPU platforms, FPGAs, and I'm sure there are a few others.
Madhatter32
macrumors 68000
Apple sells personal computers. It is not in any of the businesses you cite.
Nope. I know the difference.
When that happens we buy old hardware. You just don't realize how much companies not in the computer business path to spend money on software! It's a cost drain to them, and they're right, they don't ever make any money off of it.
When that time comes, we'll deal with it. It rarely comes though. Emulation and virtualization are pretty handy.
I've tried, but software vendors are insane with the money they want for upgrades and migration projects. I expect we'll just buy something new and shoehorn it in -- that's actually cheaper.
Tagbert
macrumors 604
Fortunately it is not emulation, it is automated recompilation. Most companies are not going to care one bit as long as their people can still do their work.
I thought he was talking about running emulated x86 on a Window-on-ARM VM, running on Apple Silicon.
Perhaps you mean Rosetta, which I understand is essentially a translation layer that runs (mostly) at installation time, with some on-demand recompilation to ARM64?
Tagbert
macrumors 604
It sounds like you have some a need to run some very specialized software with no support for Macs. That is unfortunate but if the software vendor doesn’t care about your desires, you may be stuck going back to a windows box unless Windows virtualization on Apple Silicon improves significantly in the near future.
Tagbert
macrumors 604
I thought he was talking about Rosetta which is what I was describing. This whole conversation is going in circles . It seems like the original poster’s questions have been answered and they can make their decision now. It doesn’t sound like that original poster are doing day trading with software running on Wine.
throAU
macrumors G4
Apple do not write the financial software. They provide libraries for developers to use to write apps as they see fit. The issue is the software vendor has not ported/updated their app.
And that's their prerogative; nothing to do with Apple. ?♂️
edit:
for what its worth I use Macs for things about as far from their core market as possible - I'm a system administrator / network engineer.
Last edited:
Fair point. Apple provides Rosetta that seems to work for most x86 macOS apps. As far as I know, Apple has made no claims about supporting Windows x86 on Apple Silicon now or in the future. From what I understand they haven't even made any official comment about support for Windows-on-ARM, although it appears to be technically working.
There are at least 3 major OS platform types in current usage (Linux/Unix, Windows & MacOS) on 2 major processor architectures (x86_64 & arm64) - plus the more specialised offerings in enterprise and scientific communities.
I don't have a problem accepting that not all software vendors will support all of these platforms, and doing so can be a major effort for the developers, so they are perfectly within their rights to support whatever they feel makes most sense for them.
It seems rather pointless (other than as an academic exercise or to save money) to attempt to make software that was never designed to be cross-platform run on another platform.
Maybe I could try to make my petrol car run with diesel, or convert to an EV....but why bother unless you have a unique resource that makes the effort worthwhile. (BTW there is a company that converts classic cars to use electric motors - but it is hugely expensive, and I'm sure loses some of the appeal of collecting or restoring the original engineering).
Windows computers can be had for not much money. If you need to run WinTel apps, just buy one - or use a cloud-computing instance if you don't want to have any more machines on your desk.
Shirasaki
macrumors P6
Oh, I thought it was something else. Benchmark number again...
I just throw this out there: benchmark number means nothing for most users. Experience matters. And for me, M1 isn't too much better than Intel processor when stuff is running. When it goes hot, it goes REAL hot. Just not as hot as, say, Intel processor based computer.
Also, comparing M1 vs other Intel/AMD offering isn't that relevant anymore. What we will really compare is M2/M3 vs future generations. Intel still need a few years to clean whatever 14nm++++++++++++++ mess is and move forward, while Apple's processor potential and power have yet to be fully realised.
I just throw this out there: benchmark number means nothing for most users. Experience matters. And for me, M1 isn't too much better than Intel processor when stuff is running. When it goes hot, it goes REAL hot. Just not as hot as, say, Intel processor based computer.
Also, comparing M1 vs other Intel/AMD offering isn't that relevant anymore. What we will really compare is M2/M3 vs future generations. Intel still need a few years to clean whatever 14nm++++++++++++++ mess is and move forward, while Apple's processor potential and power have yet to be fully realised.
This sounds like an emotional reaction rather than an intellectual one. What’s the problem with “emulation”? Entire industries run in virtual machines and emulated environments. The web is based on “emulation“. More so, x86 is based entirely on “emulation” - x86 binary is just a convention that CPUs translate into native microcode. In the end of the day, why do you care whether your x86 application is translated in the fly to some sort of internal code that the CPU runs or whether it is translated in the fly to ARM code the CPU runs? I would think the main concern is correctness rather than what happens under the hood.