The M1 Max's GPU is more in line with an RTX 3060 AFAIK.
Despite Apple's claims, M1 Ultra GPU is not as powerful as RTX 3090 - VideoCardz.com
No, Apple M1 GPU is not as powerful as RTX 3090 Apple may need to reconsider its M1 silicon marketing. The company was claiming that their newest chip called M1 Ultra equipped with 64-core GPU is as good as high-end desktop GPU while consuming 200W less power. As proven today, this claim...videocardz.com
M1 series especially M1 Ultra still suffer its GPU performance and it turns out Apple lied about the performance. Yes, the optimization is still a problem but people still believe that M1 series are still performing poorly. Do M1 Ultra really perform just like RTX 3090? How about now? If this continue in 2022 and later, it will be a huge problem. I'm still seeing videos of poor M1 series performance due to lack of optimizations.
Power per watt is the only advantage for M1 series so far but dont we really care more about the performance? At this point, I'm already worrying about the future of M series.
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.
M series chips are still slow... Are they gonna solve it?
- Thread starter sunny5
- Start date
- Sort by reaction score
I doubt we're going to make much more headway in this debate.
But like... your title says "M Series chips are slow" and now you're saying "the point isn't that they're slow". Which is it?
And then your first post says "dont we really care more about the performance?" and I was just responding to that, that going for highest absolute performance directly can cause issues in other areas, and the balance apple has struck I think is the right one for me. Like do you just want blind agreement?
If you want some common ground, I somewhat agree with you on the example of the Ultra. That its seemingly aggressively limiting itself given the form factor and its stated purpose (although I haven't used one so I don't know all the caveats behind those results).
Also I agreed in my very first post on this thread that Apple walked into criticism trying to compare themselves to a 3090, when the reality is their current GPUs are not really very competitive for people who'd otherwise be in the market for that type of graphics card. It doesn't mean they are bad GPUs in any way though, or they are in trouble with this hardware strategy. It just means the dummies in the marketing department are pretending to be competitive in a particular market segment which in my opinion they really aren't. If thats the main thrust of your argument, then I agree with you!
They have no point. They looked at a few graphs and extrapolated the performance under those very cherrypicked tests to every possible task with no supporting evidence, and now they're mad that the computer doesn't do a thing that absolutely no one told them it would do.
It’s also false in ad-hoc tests carefully prepared using code that’s not optimized for Apple Silicon. Anyone that has taken even the minimum amount of effort to even the playing field sees results close to what Apple’s presented. Folks looking for views or that need to create content to stay relevant on YouTube are happy to claim that Apple Silicon is slower running Intel optimized code without saying so. BUT, even saying that still highlights a truly impressive thing… that macOS can actually run code that’s been written for Intel on systems without Intel CPU’s!
If it weren’t possible to run Intel code on Apple Silicon, then they’d be forced to actually create same-for-same tests and everyone would be reporting the same results (or they wouldn’t because they won’t get enough traffic for those posts
)
Last edited:
We can dream, but clearly that didn't happen. You're still here trying to convince everyone that we're useless, slavering fanboys if we notice any area where Apple has some sort of advantage.
Most people don’t really care if their computer isn’t the fastest thing available. They just want something that is sufficiently fast for their needs. And most people are buying sub $1000 computers that simply are not in the same league as Apple’s M series chips, and they’re largely content with that. Their biggest concerns are having sufficient storage for their needs and how long before their computers become unacceptably slow to use anymore.
If people step into the Apple arena they are now getting computers that have more power and speed than most of them will ever need.
Those whining about other chips being faster than Apple’s M chips are hyper focused on a very narrow parameter of performance. They’re fixated on something most in the real world couldn’t care less about.
If people step into the Apple arena they are now getting computers that have more power and speed than most of them will ever need.
Those whining about other chips being faster than Apple’s M chips are hyper focused on a very narrow parameter of performance. They’re fixated on something most in the real world couldn’t care less about.
I have not read the thread completely. But apparently Apple has claimed something that is not true. And that has been exposed and some folks seem to have a problem with it and are talking it up with arguments that aren't even the point.
What's the point?
It is in Apple forums again and again the same
What's the point?
It is in Apple forums again and again the same
Well, if one were to think critically about the issue, would one trust Apple, whose claims likely have legal implications, as in Apple may be sued if proven false, or some YouTuber’s videos, whose primary objective is to get more views, who would one trust?I have not read the thread completely. But apparently Apple has claimed something that is not true. And that has been exposed and some folks seem to have a problem with it and are talking it up with arguments that aren't even the point.
What's the point?
It is in Apple forums again and again the same
So far, the ‘exposed’ part are based on gaming From what I can read from ’Apple forums’.
Similarly I remember videos posted that shows workloads that supported Apple’s claims. Those seemingly are swept under the rug.
Well, I supposed this is the Internet.
M1 Max compared to an i7 (that could have made it to the 27” iMac) it looks quite disappointing:
Intel Core i7 12700K vs Apple M1 Max: performance comparison
We compared Intel Core i7 12700K (3.6 GHz) against Apple M1 Max (3.2 GHz) in games and benchmarks. Find out which CPU has better performance.nanoreview.net
So which computer available for sale did they use in the test?
My neighbour’s car can do 0-60 in 4.6 seconds. My car can do 0-60 in 4.9 seconds.
My car is junk.
Uh, no, I don’t track my car or street race.
My car is junk.
Uh, no, I don’t track my car or street race.
What's actually going on is roughly this:I have not read the thread completely. But apparently Apple has claimed something that is not true. And that has been exposed and some folks seem to have a problem with it and are talking it up with arguments that aren't even the point.
What's the point?
It is in Apple forums again and again the same
* Apple posted benchmark results showing M1 Max being competitive with a particular Nvidia GPU in one particular workload. Yes, you can call this a cherry picked benchmark. Or you can call it them wanting to show what their GPU is capable of when software is properly optimized for it. Pretty standard caveat emptor stuff for vendor provided benchmarks.
* in response, various PC enthusiasts benchmarked several other things, mostly games poorly optimized for Apple's GPU, and used the results to claim Apple is lying
* Mysteriously they never try to replicate Apple's test to confirm or deny the specific claim Apple made. Nobody has actually demonstrated that Apple lied, as far as I know.
* Periodically one of the usual trolls who hang out around here starts a thread whining about how evil Apple lied
Which software did Apple use for the benchmark? The footnote of the graph says:
Mac Studio
The new Mac Studio. Supercharged by the M2 Max and M2 Ultra chip. Extensive connectivity. Stunningly compact.
www.apple.com
Has Apple explained how to replicate the graph? Has anyone been able to replicate the graph?
Tech Youtubers have tried to replicate the chart by doing various benchmarks because Apple has not disclosed how it did the benchmarks.
You're not doing the same things with an RTX 3090 that you would do with an M1 Ultra's GPU. It's Apples and Oranges.
It is quite obvious they tested content creation applications like Adobe photoshop, illustrator, Premitre, Lightroom, Da Vinci resolve or Affinity suite of applications. Any of those industry standard applications will show the ultra equivalent or exceeds the rtx3090.Which software did Apple use for the benchmark? The footnote of the graph says:
Mac Studio
The new Mac Studio. Supercharged by the M2 Max and M2 Ultra chip. Extensive connectivity. Stunningly compact.www.apple.com
Has Apple explained how to replicate the graph? Has anyone been able to replicate the graph?
Tech Youtubers have tried to replicate the chart by doing various benchmarks because Apple has not disclosed how it did the benchmarks.
For pure graphics probably gfxbench which is native and optimised for Metal and is within 90% of the rtx3090.
Fyi as a side not - I have no idea with the obsession with cinebench. The application is hardly used and is known to be not optimized for Apple Silicon. It is almost always brought up when trying to say pcs are superior to Apple Silicon. I’d rather test applications that are hugely popular and widely used.
“Slow”
This must be some strange definition of the word slow that I wasn’t previously aware of.
This must be some strange definition of the word slow that I wasn’t previously aware of.
I'm glad they have the current Ultra, that Apple allowed the low end Mac Pro chip to be used in such a small case, and that it is as slow as it is.
Cannot wait for people to get their hands on the Quad version of the Ultra when the Mac Pro arrives. It should be classified as "Not as Slow"
You’re right, and its not a strange definition at all. Slow and fast are relative words. A turtle its fast if compared to a snail. And an F1 car its slow if compared to a fighter aircraft.
What can't you do with an RTX 3090 that you can do with an M1 Ultra? Video/photo editing? 3D rendering? Gaming? Machine learning?
When it suits Apple, it explicitly says what software it uses for the benchmark. For example, Apple describes what software it used for the comparison between its new hardware and its older hardware.
These are some of the footnotes for the comparison between Mac Studio and Mac Pro.
However, Apple tends to use undisclosed benchmarks when comparing its hardware to that of other companies.
Efficiency. The RTX 3090 requires far too much power to do what it does. There are only select few tasks where it outshines the M1 Ultra.
It's not that there are things that you can't do with one that you can do with the other. It's that the GPUs on the M1 Ultra/Max/Pro are not engineered for gaming first and foremost in the way that the GeForce RTX GPUs are. There's certainly overlap between the two, but each one has its specialty. I think a better comparison for Apple/NVIDIA to be making when it comes to GPUs is to compare the Apple Silicon GPUs to the NVIDIA Quadros. Because those are at least geared toward the same intended applications.
Then again, during the PowerPC and Intel eras Apple was always more trying to load in gaming GPUs for workstation tasks and then claiming workstation performance, when they could've just used workstation line GPUs. Moot now that Apple is the one making GPUs for Apple Silicon Macs.
Of course they are. In fact, in comparative terms M1 does better against Nvidia in gaming than in pure compute.
I'm not sure the comparison between Nvidia's workstation GPUs and Apple's Ultra GPU is fair because Nvidia's GPUs are more expensive.
The RTX A6000, Nvidia's most powerful workstation GPU, is as powerful as the RTX 3090 Ti and consumes 50% less power.
NVIDIA RTX A6000 Specs
NVIDIA GA102, 1800 MHz, 10752 Cores, 336 TMUs, 112 ROPs, 49152 MB GDDR6, 2000 MHz, 384 bit
NVIDIA GeForce RTX 3090 Ti Specs
NVIDIA GA102, 1860 MHz, 10752 Cores, 336 TMUs, 112 ROPs, 24576 MB GDDR6X, 1313 MHz, 384 bit
How can it consume less power? It’s literally the same chip.I'm not sure the comparison between Nvidia's workstation GPUs and Apple's Ultra GPU is fair because Nvidia's GPUs are more expensive.
The RTX A6000, Nvidia's most powerful workstation GPU, is as powerful as the RTX 3090 Ti and consumes 50% less power.
NVIDIA RTX A6000 Specs
NVIDIA GA102, 1800 MHz, 10752 Cores, 336 TMUs, 112 ROPs, 49152 MB GDDR6, 2000 MHz, 384 bitwww.techpowerup.com
NVIDIA GeForce RTX 3090 Ti Specs
NVIDIA GA102, 1860 MHz, 10752 Cores, 336 TMUs, 112 ROPs, 24576 MB GDDR6X, 1313 MHz, 384 bit
It is even stranger. The RTX A6000 consumes less power than the RTX 3090.
Igor's Lab did performance tests with these two GPUs.
What can the full Ampere do? NVIDIA Quadro RTX A6000 with 48 GB VRAM as a 8K and 5K card in a first gaming test - pure decadence and a small victory against the GeForce RTX 3090 | igor´sLAB
Of course, what I’m doing today is just a first taste of what’s to come later in full glory. But large workloads also take time to prepare, because it’s supposed to add real value. And that’s where I’…
www.igorslab.de
It is even stranger. The RTX A6000 consumes less power than the RTX 3090.
Igor's Lab did performance tests with these two GPUs.
What can the full Ampere do? NVIDIA Quadro RTX A6000 with 48 GB VRAM as a 8K and 5K card in a first gaming test - pure decadence and a small victory against the GeForce RTX 3090 | igor´sLAB
Of course, what I’m doing today is just a first taste of what’s to come later in full glory. But large workloads also take time to prepare, because it’s supposed to add real value. And that’s where I’…
Had a quick look, the 3090 uses much hotter GDDR6X, so that's probably where the difference comes from. Plus, probably higher binned chips on the A6000 side.