Any code which is *not* linking to a Windows API and hence could be redirected to Wine's implementation is running unaltered on the CPU. No Intel CPU, no execution of that code.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
The final blow for real gaming on the Mac, or the beginning of a new era?
- Thread starter Frankied22
- Start date
- Sort by reaction score
Of course it is altered. It's altered by Rosetta. How do you think Rosetta works? It intercepts x86-64 binary and translates it into ARM64 binary.
Anyway, I feel that we have derailed this thread enough. Let's just wait and see at this point. The ARM Macs will come out soon enough and then one can just try it out.
I think the fact that this thread has gotten so many replies shows that there are still those of us who want to see the Mac as a viable gaming platform. As I said in my OP, I think this will take a few years. Apple Silicon will prove to be impressive in both CPU and GPU, windows will probably move to ARM, and more game studios will start to develop for ARM and get their games out to a bigger user base. I could see a lot of smaller indie games releasing on both Mac and iPad. There are already a lot of hit indie games with iPad versions.
I know. But the only company using Wine regularly as basis for their Mac ports has been out of business since 2015 and has only ported a small number of games.
The only two remaining games for Mac still using Wine in one form or the other and having survived the 32-bit culling are Eve Online and Final Fantasy XIV.
Have you personally tried flight simulators via cloud? I would think the base sim would run, but the minute flight sim “pilots” want to use plugins, or more advanced peripherals such as head tracking, simulated radio panels, multiple screens I would think remote cloud computing would not be compatible.....
Microsoft‘s upcoming flight Sim 2020 is the only reason I would need a powerful pc, otherwise a cheap win 10 laptop would work for any apps that i currently use which don’t have a macos equivalent yet. If cloud computing can interface with local usb devices, it would be a great option for me.
Those two posts pretty much sum up my opinion.
With the exception of maybe StarCraft on occasion, most of my gaming is relegated to a dedicated console, and i'm fine with that setup.
I'm not holding my breath, but Apple might have a small ace-up-its-sleeve with the marketshare of iOS, and how easily those apps can be brought over to the Mac. In theory, superficially speaking, if a dev makes a game for iOS they could add support for Mac controls(keyboard and mouse). I wouldn't expect AAA third-party stuff.... but the mobile and indie stuff, why not?
For example, "World of Goo".
iOS/iPadOS 14 bring KB+mouse support to games, along with greatly enhanced Dualshock/Xbox controller support. A lot of the games coming to macOS from iOS will already support desktop controls.
Last edited:
The most popular games are not the big blockbuster releases anymore. The future (and now) is cross-platform online multiplayer games. For example Fortnite, you have iPad, console, PC, and Android players on the same map. Sure, it’s a pain in the ass for programmers. But it is the most lucrative way to develop games now. and this is great news for gaming on ARM Mac
For apple game developers I don't even think it will be that bad.
Code your game for cross-platform between Apple Silicon Mac and iOS using Metal and Swift; with only minor tweaks it will also run on iPhone when the hardware catches up the following year or two.
People think this is going to be a problem - there are plenty of games on iOS devices, making Mac the same architecture and essentially the same development framework is a WIN, not a loss - x86/x64 Mac was a minority platform, collectively all the Apple Silicon devices are larger than any other segment in the market.
Even an iPad from 2 years ago is more capable than a Nintendo Switch; the only thing that has been holding back proper games from these platforms is controller support - and its been out for a little while now.
That said, I really do think Apple need to put out a first party controller, or at least more heavily promote the sony/xbox options in store or on their online store.
If Apple really wants, they can integrate a x86 coprocessor. It could be initially outsourced from AMD/Intel, and then they could manufacture their own coprocessor that outputs less power. It would only be called when legacy X86 code is executed.
Since we're at a time that computer components have gotten so small, I don't see how they could use price or space to object to that. Our current time is very different from when they made the transition from Power PC to Intel, and it requires a different solution.
If you think you would never be able to have decent x86 gaming performance at a tiny space, think again, because the GPD Win Max can even run software as heavy as Witcher 3 (although in lower resolution).
Granted, this has the downside of making the architecture a bit more complex, but processors are so small these days that would it get that complex, or that expensive? Plus, the coprocessor could share all the hardware resources (e.g, video card), and since x86 is running at native level, you get native speed.
Note that Apple has done that before with Apple III through a PCI card. Apple III ran Apple II code at a better resolution and faster speed than the original Apple II hardware. And they could make it optional (e.g, could be added or removed at purchase time) if you don't want that feature. Or they could make it pluggable.
IMHO Apple could only really take a share of the X86 market with this feature. X86 is not just cheaper because of the hardware itself, but also because the legacy compatibility allows you to save a lot of money. Instead of buying 2-3 pieces of hardware to play a couple of games, you only need one computer. And it seems Intel is already on the move to work on hardware acceleration, so I don't think Intel (and much less AMD) will stand still.
Since we're at a time that computer components have gotten so small, I don't see how they could use price or space to object to that. Our current time is very different from when they made the transition from Power PC to Intel, and it requires a different solution.
If you think you would never be able to have decent x86 gaming performance at a tiny space, think again, because the GPD Win Max can even run software as heavy as Witcher 3 (although in lower resolution).
Granted, this has the downside of making the architecture a bit more complex, but processors are so small these days that would it get that complex, or that expensive? Plus, the coprocessor could share all the hardware resources (e.g, video card), and since x86 is running at native level, you get native speed.
Note that Apple has done that before with Apple III through a PCI card. Apple III ran Apple II code at a better resolution and faster speed than the original Apple II hardware. And they could make it optional (e.g, could be added or removed at purchase time) if you don't want that feature. Or they could make it pluggable.
IMHO Apple could only really take a share of the X86 market with this feature. X86 is not just cheaper because of the hardware itself, but also because the legacy compatibility allows you to save a lot of money. Instead of buying 2-3 pieces of hardware to play a couple of games, you only need one computer. And it seems Intel is already on the move to work on hardware acceleration, so I don't think Intel (and much less AMD) will stand still.
I'm no expert in the field of microarchitecture, but it seems to me that the chiplet design of AMD's Ryzen CPUs would be well-suited for this. Instead of a full-bore x86 CPU, they could integrate 1-2 Ryzen chiplets into the A-series SoC die. Apple and AMD even share the same fab provider and process (TSMC 7nm/5nm), so it could presumably even all be done in the same fab plant.
They could buy them in but there's no way they'd bother making their own at this stage. They'd need to licence the instruction set from Intel anyway and I don't see that happening at all. The only licence it to AMD because they were forced to.
Last edited:
Very unlikely. Apple want developers and customers to migrate to ARM. Adding x86 cores would defeat that purpose.
That’s exactly what i was saying.
For you hardcore gamers, let’s say apple stays with intel forever, did you or would you consider any Mac computers as your ‘pilot’ rig?
on the other hand, for us ’casual’, ‘basic’ gamers, this cloud gaming probably will do good (I haven’t tried every game of course), and this will run on any Mac/iOS device, with the iOS 14 supporting mouse/keyboard gaming. And we pay only ~$100-ish/year for that, which is also much cheaper than a dedicated gaming setup. I’d say there‘s never a better time For apple gaming.
Wasted power, wasted motherboard space, increased complexity, additional heat, etc. for everything that isn't native applications. Which should be the vast majority of stuff moving forward.
This will never, ever happen. Apple aren't building these machines for yesterday.
Rosetta2 will be fast enough, and the performance difference if one does exist can be made up by making the ARM chip faster, Rosetta2 more effective and use of natively compiled frameworks.
i.e., the power/space/heat/etc. that is used for some crappy duct-taped on x86 core is far better put to use to make the native Apple Silicon faster so that the translation performance improves.
Never mind the design issues of memory/bus contention between the native and x86 processors (both processors can't use the same memory without locks or some other scheme to prevent one processor overwriting memory and invalidating the other's internal cache as just ONE example of why this is difficult), additional space for motherboard traces, heat problems, etc., etc.
It's very easy to sit back and say "just add another cpu to the machine!" but in reality the tradeoffs to that are significant, and it would suck. You would not get seamless full speed x86 performance anyway, due to the above issues.
You'd just end up with a large, hot, heavy machine that was worse at running native software and sucked at x86 software anyway due to all the trade-offs to shoe-horn in an x86 chip and associated hardware.
The translation method Apple has decided on (again, for the second/third time around) would actually most likely be a LOT faster, unless you included basically two entire systems (I.e., two sets of RAM, two system buses, etc. - which is a LOT of hardware that will be entirely useless for native software) in one, as invalidating the cache on a modern CPU makes performance drop off a cliff.
Try turning off all of your CPU's L1/L2/L3 caches on a PC if the EFI/BIOS supports it and see for yourself.
Last edited:
I think Apple might be further than we think after reading this slide
View attachment 931653
Last edited:
They could if they wanted to, sure.
They won't because it would suck as per my post above, and they're catering to maybe 1% of the Mac market, which is just not worth it.
You think the thermals, power consumption and memory bus sharing is all free? By trying to be everything you will suck at everything due to spreading the budget for space, power and heat far too thinly, never mind the additional cost.
[automerge]1594163951[/automerge]
Yeah they sucked too. By 1995 the pentium 133s, etc. were out and many times faster than a 486. Never mind that 90% plus of all sales these days are portables where heat/battery/space and price all matter.
Spending the additional thermal, power, space and dollar budget on dead-end Mac hardware that will not be used inside of 2 years for busted x86 hardware is dumb. I haven't even gone into the software side of making it work yet either - and Apple's small software team has a lot more useful things to be spending their resources on.
macOS and iOS are buggy enough as it is.
If you want PC compatible hardware for games, buy a PC (or maybe a console). It will be cheaper than the additional Mac hardware and far better at doing that job.
Last edited:
How did you become a mod?
It was running on an A12Z ARM chip, an iPad chip. Not desktop, please get educated first then complain. There will be dedicated Mac chips soon...