And what have you said confirms there is a change in demand? Nothing. So maybe that's why you're not sure where the debate is.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Metal Developer Tools on Windows Offer New Gaming Opportunities
- Thread starter MacRumors
- Start date
- Sort by reaction score
And what have you said confirms there is a change in demand? Nothing. So maybe that's why you're not sure where the debate is.
What interesting about the discussion in this thread is that they feel like déjà vu.
Every once in a while there are these discussions about gaming and Macs(and just to be clear I'm talking about high end gaming, 60$/game level stuff), this illogical optimism about Macs becoming relevant as a high end gaming platform.
Of course Vulkan is actually getting more support and starting to become the preferred alternative API to DirectX. In the end it makes sense, Vulkan is supported on both Linux and Windows so the potential user base is much larger than Metal anyway.
Apple hasn't done anything relevant to change it's situation regarding high end gaming and the move to ARM surely won't help them because they will still lack the necessary user base to peak the interest of big gaming studios.
In the end the only hope for Mac is if apple throws around a lot of money and invests in ports or exclusives.
Last edited:
Almost 0 chance to happen but it is possible that GTA VI will support Vulkan(just like Red Dead Redemption 2 does).
I am confused, one guys says Metal+ARM will make games more possible on MacOS, another says its not going to happen.
It will come down to Engine support I think. Games that use UE4/5 and Unity will likely be easier to port than games that use other engines (like REDEngine or IDtech 6/7). If the engine doesn't support Metal it is likely the devs may not bother if there isn't an existing userbase.
The only way I see AAA games being ported to Apple Silicon is if Apple actually pays the devs to do it. Apple needs to create the market and demand. They are in the mobile game market but they have a very enticing platform for AAA games as well. Especially if they focus on the GPU on their SOC's.
If they are serious about gaming (Which they should be, since it's as big as the Film/TV industry) then they should reach out to the top 3 or 5 game studios and get some AAA titles over to the Apple platform. Imagine if Apple works with EA and gets some sports titles like Madden and FIFA (Not the crappy mobile versions)? Or even the Star Wars games? It's all about money. If one studio makes a killing, they will all start supporting Apple Silicon.
If they are serious about gaming (Which they should be, since it's as big as the Film/TV industry) then they should reach out to the top 3 or 5 game studios and get some AAA titles over to the Apple platform. Imagine if Apple works with EA and gets some sports titles like Madden and FIFA (Not the crappy mobile versions)? Or even the Star Wars games? It's all about money. If one studio makes a killing, they will all start supporting Apple Silicon.
Well the talk in the past was that Metal, when is going to launch is going to "make games more possible on MacOS". That obviously hasn't happened so how is ARM going to change anything when one of the biggest problems is the small user base?
The switch to ARM will be fatal to Mac gaming on boot camp (which I wouldn't call "Mac gaming"), but I failed to see how this would be detrimental to macOS gaming.
When boot camp was announced, people predicted the end of OS X gaming, as boot camp would presumably take over. That prediction was sensible, yet it didn't materialise.
Now, I see no reason why macOS gaming would suffer from this transition. Macs will mostly have better GPUs (bye bye intel), Metal will improve (as it is tailored for Apple GPUs) and macOS will be able to run iPhone and iPad games.
[automerge]1594883780[/automerge]
Games using modern engines (namely Unity and UE) wouldn't have macOS versions had Apple stuck to openGL.
When boot camp was announced, people predicted the end of OS X gaming, as boot camp would presumably take over. That prediction was sensible, yet it didn't materialise.
Now, I see no reason why macOS gaming would suffer from this transition. Macs will mostly have better GPUs (bye bye intel), Metal will improve (as it is tailored for Apple GPUs) and macOS will be able to run iPhone and iPad games.
[automerge]1594883780[/automerge]
Huh?
Games using modern engines (namely Unity and UE) wouldn't have macOS versions had Apple stuck to openGL.
Last edited:
OpnGL has been phased out by Vulkan in gaming in case you didn't know.Huh?
Games using modern engines (namely Unity and UE) wouldn't have macOS versions had Apple stuck to openGL.
AAA gaming, we are talking about AAA gaming not gaming in general which also includes 1-3$ mobile games.
Anyway picture this: a very small user base(intel Mac swill remain the majority of Mac computers for years to come) of computers that are not necessary the best choice for gaming. How is that going to pique the interest of big gaming studios that make AAA games? For example Rockstar Games.
In the end "ARM" is for the CPU but in order to enjoy high end gaming you need a powerful GPU(with support for VRR, Ray Tracing and other modern gaming features) above all else. And also when it comes to gaming the price of the overall system is very important. A cheaper Windows PC or laptop that also has the benefit of a huge AAA gaming library will always win vs a Mac PC that's quite noncompetitive by comparison.
The only hope for AAA gaming on Macs is if Apple pays developers to port their games.
Last edited:
So? The argument was that Metal didn't help Mac gaming. The relevant comparison is "Metal" vs. "what we had before Metal", not "Metal" vs "some hypothetical scenario".
Well it didn't as the AAA gaming situation on MacOS didn't improve after the adoption of Metal.
That doesn't change my argument. I fail to see how the switch to AMR should be detrimental to gaming on macOS. This includes both AAA and casual gaming.
Why would developers have less incentive to port their games after the switch? It might be the case if the Mac market share plummeted, but why would it?
What I see, though, is the absence of bootcamp as some new incentive to create macOS versions.
Catalina was the most detrimental thing to happen to macOS gaming in recent times. ARM Macs, not so much. They will run 64-bit intel games quite well under Rosetta. Shadow of the Tomb Raider is running way better on the A12Z under Rosetta than on any intel GPU (at least under Windows).
[automerge]1594898588[/automerge]
If you mean that the percentage of AAA games ported to the Mac didn't increase after Metal, you may well be right (though we don't have these numbers). The 3D API has indeed little to do with that. What counts is the market share of the target platform.
However, Metal allowed quite a few AAA games and engines to be ported to macOS, as openGL 4.1 lacked many features, namely compute shaders.
Last edited:
Good point, I remember that. But the upcoming Xbox and PS5 are based on RISC shouldn't that maybe make things easier?
The fact that you fail to see something doesn't mean you are on to something.
I said it wouldn't help AAA gaming on MacOS and you don't have to see it to be true.
Because of the MUCH SMALLER USER BASE. Its quite simple actually. Even the current Intel Mac OS user base which is basically 15 years old, doesn't pique the interest of big gaming studios so the vast majority of PC AAA games don't get a MacOS port. Or you would disagree with this simple fact?
ARM Macs won't really do anything to improve the AAA gaming situation on MacOS, thta's my point. And you do realize you are talking about a single 2018 game?
That's questionable, we didn't see the resolution, settings(to me it looked like it was mostly on low settings) and most importantly the FPS. They also only showed a simple static scene with no NPCs and not much happening so the best case scenario for a game in terms of fps.
Anyway Intel's next generation iGPUs are rumored to bring 40-50% improvements in FPS in games so it's not like intel's iGPU performance will stay the same as it is today. The idea is thta all gaming laptops and gaming PC's have dedicated GPUs so even if Apple's integrated GPU will somehow be faster than Intel's integrated GPU it won't mean much in the end if somebody interested in gaming can just buy a Windows laptop with a dedicated GPU at the same price(or cheaper) he would pay for a Macbook with an integrated GPU.
[automerge]1594899227[/automerge]
NO.
The upcoming Xbox and PS5 will use AMD x86 Ryzen CPU and AMD RDNA GPUs and they will be able to run AAA games at 4K/60fps.
[automerge]1594899415[/automerge]
Of course I meant that the percentage of AAA games ported to the Mac didn't increase after Metal. There was a spike for 1 years after the introduction of Metal but the numbers went down after than and, the end.
Last edited:
Food for thought... Nintendo has been using RISC processors. One would wonder why studios didn’t port a lot of their games from their PS/PC/Xbox catalogs
Power difference between them has been the largest impediment. Nintendo stopped chasing raw power (openly) after the N64.
My point is that the ARM switch should not be detrimental to macOS gaming. I never said it would "help AAA gaming on MacOS". I wasn't initially replying to your post, btw.
Much smaller than what? The current Mac user base?
Even written in all caps, I don't agree that the user base will be much smaller since developers can target both intel Mac and ARM macs with universal binaries. So the user base will remain "all Macs". Did the Intel switch see a decrease in macOS gaming? If it didn't, why do you expect to see a decrease now? Universal Binaries are easier to build than during the Intel switch, due to the fact that basically all devs are now on Xcode and because of similar endianness between ISAs.
Also consider that developers won't base their decisions on the initial user base. Do you expect the Mac market share to drop in the long term?
The resolution was 1080p, as said in the keynote, and the mean frame rate was >30, since we see a new frame in the game at every frame of the video stream (except for a couple of frames when she hits the water). We may be talking ~40 fps on average, but no one knows.
Even 30 fps is way better than what intel GPUs can do on the benchmark tool: https://www.notebookcheck.biz/Shado...-pour-PC-portables-et-de-bureau.333354.0.html
Ok, the section that was demoed in the keynote was likely less demanding than the benchmark tool, but consider that the best intel GPU can only generate 27 fps on average at 720p low.
In general, there's no question that intel GPUs are not very good. Maybe they'll improve in the future, but I highly doubt they'll beat Apple in perf/W. They'll be using intel's inferior lithography and slow interface with the system RAM (via PCIe), while Apple GPUs will most likely use the best TSMC's lithography and and have direct access to system memory (i.e., at RAM speed).
Last edited:
I don’t think what Apple showed should be directly compared to the in game benchmark (where the Intel numbers come from).
One would need to compare performance in the section shown in the keynote to the benchmark tool. I own the game, but I haven't yet reached that section.
All the videos I've seen of the game running on intel GPUs look pretty bad and choppy, regardless of the section that is played. I haven't seen a video using the most powerful intel GPUs though.
I got sucked into playing Control on my PC so I haven't gotten around to finishing RotTR yet.
Having a RISC processor will have no impact on the portability of game code. The PS3 had a RISC processor, but the problem is that there was not enough R&D going into keeping that processor competitive. The actual fact is that Intel's CISC x86 processor ran off the rails a long while back because RISC processors were winning - to compensate for this and to keep the x86 competitive Intel came up with a fix that kept them in the game (as long as their fabs were at or near the top in the world - something that is not the case right now). The fix was to swap out the CISC CPU and replace it with a RISC CPU, then add another layer using microcode translation from CISC operations to RISC microoperations using hardware translation. I suspect this is actually the real reason why Intel has been unable to create an energy efficient x86 CPU to compete with ARM in the devices. The fact that Apple A series chips do not have this extra technology means the cores themselves are smaller and likely will be able to fit more cores on their CPU. Apple also has been able to optimize their CPU so that for single thread use it also is very competitive (and that is within a very thermally restrictive environment). All things being equal, Apple should be able to design a CPU/GPU in a similar fashion to the AMD processors that are used in game consoles - but with more room left. Apple will also be using the 5nm fab to fab the Mac processors - which will only increase the efficiency / power. Apple probably has more chip designers than AMD (at around 1,000), and a very good team of them in place. The A series chip and the x86 chips are both little endian (Power PC is big endian) [little-endian vs big endian is how integers are stored - they are the reverse of each other -- which should remove one annoyance when it comes to cross-compiling. Most games are written (mostly) using C++ or other high level languages - assembly language would likely not be used and if it was it would be a rather small part of it and easy to add conditional logic for specific processor. This should not be necessary in the vast majority of cases though. Apple has a lot of options available to them because their CPUs are much more efficient for the same power than x86 chips - it means you can pack more power in before reaching the thermal limits. Apple can even go the way of AMD chips with chiplet design options with specialized co-processors that will add more performance for applications using the Metal interface. I use to write x86 (and various other processors) assembly code many decades ago and the architecture (which is backwards compatible) was already archaic back then. It was my least favourite processor to write assembly code for. Intel has had success primarily in spite of their processor design. They became the standard originally because of a decision IBM made, and as long as Intel had a lead in their manufacturing they could effectively keep the fact that their x86 instruction set was archaic a non-factor for desktop and laptop computers. IMHO, it would have been better for the entire marketplace if x86 were effectively retired 30 years ago. I look forward to a day where Wintel monopoly on Windows processors in quality computers comes to an end.
And let's not forget these results : https://www.anandtech.com/show/13661/the-2018-apple-ipad-pro-11-inch-review/6
On the GFXBench test (at high settings, because the "normal" settings are CPU-bound), the A12X puts all the competition to shame, including the Ryzen part, which uses an AMD RX vega 10 rated at 1.6 TFlops and running the test with Vulkan. I suppose the AMD part was not running at boost clock speed (1300 MHz) though.
On the GFXBench test (at high settings, because the "normal" settings are CPU-bound), the A12X puts all the competition to shame, including the Ryzen part, which uses an AMD RX vega 10 rated at 1.6 TFlops and running the test with Vulkan. I suppose the AMD part was not running at boost clock speed (1300 MHz) though.
Does anyone know if Metal is using half precision? The article wasn't sure which could explain the speed differences.