AMD's offering will give you full Windows compatibility, which is why it makes sense to go with a battery workaround (e.g, a powerbank). All the USB accessories are also 100% compatible with it.
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.
Why Intel and AMD don't make chips like the M2 Max and M2 Ultra
- Thread starter sack_peak
- Start date
- Sort by reaction score
AMD's offering will give you full Windows compatibility, which is why it makes sense to go with a battery workaround (e.g, a powerbank). All the USB accessories are also 100% compatible with it.
I have it plugged at my 27'' 4K monitor and it runs beautifully. You don't have to run it at 1080p.
Why would they need to since Apple Silicon is not even close to AMD and Intel in terms of performance.
Performance-wise, Apple Silicon chips are pretty good. But Apple wants you to do computing their way, and this interferes with what they can do.
For example: the iPad pro can run the Wii dead cold. But since Apple is against emulation, you won't find an easy way to do it. If you dare use an alternate store, Apple will periodically block your emulator.
Of course, I only mentioned emulators, but there are many other applications the iPad will be restricted to. For example, you can't write your own custom drivers to the iPad unless Apple gives their blessing (and you pay $$$).
So, it's not really the hardware, but the software (and Apple).
I wouldn’t say Amd and Intel are that far behind performance wise, but I get your point.
Hmm I think Apple got a lot of value out of the M1, considering the number of units they shipped across Mac and iPad product ranges. That’s a big platform to spread development costs on, which ultimately makes the chips cheaper in total cost to the company.
It’d be interesting to see a breakdown of exactly what the per-unit costs are of Apple Silicon, and a comparison with what Intel used to charge. I suspect Apple Silicon comes out a clear winner, because of the sheer scale of the operation. Where they used to use a range of Intel chips over the years they now use just a handful of their own.
Intel and AMD have a heavily subdivided product range which they sell into the general market, with a lot of different applications.
It’d be interesting to see a breakdown of exactly what the per-unit costs are of Apple Silicon, and a comparison with what Intel used to charge. I suspect Apple Silicon comes out a clear winner, because of the sheer scale of the operation. Where they used to use a range of Intel chips over the years they now use just a handful of their own.
Intel and AMD have a heavily subdivided product range which they sell into the general market, with a lot of different applications.
You mean the A17 which gets similar single core performance to the 12900k?
That “nowhere near”?
More like the difference between the 13900KS and the A17 -- that nowhere near. And then the intel/AMD machine can get a decent GPU as well...
Are you seriously considering the fact that Intels fastest enthusiast-level desktop CPU is barely faster than Apples smartphone as some sort win for Intel? Wow. That’s a delusion on an entirely different level.
People are getting 3000 gb6 points On the A17. No point pretending it isn’t practically the same in single core.
It's not barely, it's a LOT faster. After all, it wasn't designed with the same thermal/power limits in mind. The delusion is not at all mine.
In what way is it a LOT faster? Can you elaborate please. I would appreciate if you could reference something objective.
And they're getting more than that (~4000) out of the 13900KS, *last* years chip -- and then there's multicore *and* you can add a real GPU.
https://browser.geekbench.com/v6/cpu/1535093 (That isn't the highest, though it's close)
lol. That is not the average score. Overclocked or fake.And they're getting more than that (~4000) out of the 13900KS, *last* years chip -- and then there's multicore *and* you can add a real GPU.
https://browser.geekbench.com/v6/cpu/1535093 (That isn't the highest, though it's close)
It's easy to get high single core performance when you use a small number of cores. This does not work on desktops very well hence no proper Mac Pro with M-chip. Nobody really cares about the single core performance on desktops (and even laptops) let alone the servers anymore.
It's geekbench's problem then, but then there's no proof of your side of the agreement either. I'd say it's probably overclocked, but you can do that with the intel chip -- unlike the A17.
And it ignores multi core, GPU and thermal differences, which are pretty important for performance.
I’m not sure what you mean. You mentioned a solitary geekbench score. I mentioned the average score of the KS. Proving your score is correct is very much your problem I’m afraid.
No idea why you’d mention multi core when comparing a phone to a desktop. Compare the multicore score to a Max or an Ultra.
I only picked that one because it was the middle of the geekbench browser page.
No, I never said it was the average score. I don't even know how to use the geekbench site to find the average score. I truly hate benchmark discussions -- I only jumped in because of what leman said not conforming to what I've seen of the intel and Apple processors. I know the 13900KS is the CPU of a MUCH MUCH faster machine than an iPhone. The iPhone would start throttling almost immediately and that blows away any single core benchmark argument.
I wasn't the one that said the A17 was a fast as a 12900K (an ancient chip), was I. It's a stupid argument, the whole comparison doesn't make sense.
what?
I said my post was the average score.
How do you know that? The evidence says otherwise on a core by core comparison.
No it doesn’t. Most tasks can be completed before throttling occurs. Throttling which occurs because the chip is in a phone, not because of the chip, which is the subject we are discussing.
It absolutely does. You just don’t like the conclusion
A stock A17 Pro gets around 2930 points in GB6 single-core.
A stock 13900KS gets around 3100 points in GB6 single-core.
That's a difference of ~ 5%. How is this a LOT faster?
What's more, Intel's fastest mobile CPU — 13980HX — is slower than Apples smartphone, despite a massive difference in design power.
Yes, it's 5-6x difference in power consumption for 5% difference. What a great win for Intel. I mean, I understand that you have your reasons to prefer x86, but I just don't see how you can keep a straight face claiming that x86 has a massive performance advantage over Apple. Yes, they have a small advantage — but they pay dearly for it. While Apple is achieving this performance using less power than the most modest of mobile x86 CPUs on the market.
And they're getting more than that (~4000) out of the 13900KS, *last* years chip -- and then there's multicore *and* you can add a real GPU.
https://browser.geekbench.com/v6/cpu/1535093 (That isn't the highest, though it's close)
That is a massively overclocked CPU which at 7Ghz (according to GB logs at least), where a top expected frequency for a 13900KS is ~ 5.8Ghz. I wonder how much power that processor draws to get this result, 60-70 watts for single-core?
Last edited:
Funny, TSMC 3nm vs Intel 7 = TSMC 7nm is already a joke. If you really wanna compare, Intel needs to make CPU which is close or similar to TSMC 3nm. Otherwise, you are totally ignoring the nm advantages. Beside, Apple Silicon can't go beyond 13900K such as threadripper series.
What what, the meaning is pretty clear!
But I never did.
BS. You should be looking at a machine to machine comparison.
LOLOLOL! Try running a game, or a video conversion, or heck, even a macro in a spreadsheet.
You are so full of it -- I'm out..
Because there's more cores there, on both the A17 and the intel chip, and ignoring them because you thing it's close to the same thing is ignoring the bigger picture. You're argument is irrelevant to a machine running, whether it's a phone or a full fledged computer.
Anyway, I'm out of this argument you guys are getting way to unrealistic to me. I use computers and phones, I don't only use single cores so single core benchmarks mean nothing to me.