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.
No mention of MHz/GHz
- Thread starter lostngone
- Start date
- Sort by reaction score
Why would they? What would you compare it to?
It's a whole different instruction set and they don't want people to compare the low clock speeds of the M1 chip (prob tops at 3Ghz) to the high clock speed of x86 Intel processors (tops at 4.5-4.8Ghz), because people think "higher is better".
It's a whole different instruction set and they don't want people to compare the low clock speeds of the M1 chip (prob tops at 3Ghz) to the high clock speed of x86 Intel processors (tops at 4.5-4.8Ghz), because people think "higher is better".
Processor clock speed isn't a meaningful measure unless you're comparing two identical chips. They don't quote clock speed on iPhones, iPads, Watches, HomePod, or AppleTV. I assume they are doing the same with Macs moving forward.
They didn't give a single useful performance metric. We know only slightly more now than we did before the event started.
It's what @bobmans says. Comparing frequency is not advantageous for Apple, since their chips are clocked much lower. But they do significantly more work per clock. Focusing on performance benefits is the wining marketing strategy.
Like the speed difference between the Air, Pro and Mini?
Yes, I get this is a different architecture from x86 and that they won’t be equal. There are even two different types of cores in this CPU but it would still be nice to have a baseline or something rather than a completely arbitrary X times faster.
Review units are already out, and end users will start receiving the base model Air and Pro a week from today, so we'll know soon enough.
Yeah, I get that, but it's annoying that they spoke for 45 minutes and basically said nothing. From the fineprint on their site, it looks like that 2.8x speed is compared to a 1.7GHz i7 production model MBP (which as far as I can tell appears to be a 2019 13" with an 8th gen i7-8557U).
You are perfectly right, but the simple truth is that it doesn't make much sense for Apple to expose this information on their marketing page. It won't help them sell more of these units (actually, the opposite might be true). We will have figures soon enough through reviewers.
Another thing to consider is that the nature of these SoC's make spec comparison more complicated. It is a fluid range of frequencies and capabilities really, depending on what task the system needs to do at the moment.
Agreed. Without the actual GHz rating then we're missing out on an ability to compare like for like.
You can't compare x86 clocks to Apple Silicon clocks, different architecture.
Sure you can. Apple spent years doing it with PowerPC and Intel. Now it doesn't suit them they don't publish the speeds so we don't even have an opportunity to compare the ARM Mini to the ARM MBP. Somebody will need to smash these things open with a sledgehammer to see if the chips are the same or different.
Very slick marketing on Apple's part in all this. All this 2x, 3x, 6x faster... compared to which chips and graphics cards exactly? Intel 8th gen from 2017?
Clock speed would also be useful to compare the MacBook Air to the pro, or to the iPad.
Clock speed would also be useful to compare the MacBook Air to the pro, or to the iPad.
Apple also made sure to include pipeline difference to make sense of the MHz war happening. Pentium 4 chips had excessive pipelines to achieve those high clocks.
Yes, that’s the model it is replacing. If so, it does seem fairly impressive. I expect the MacBook Air to perform somewhat better than the iPad Air, and the MacBook Pro to sustain power longer because it has a fan.
Given the performance numbers I estimate 4GHz 'turbo'. Still lots of headroom to get to 4.9GHz for a future Mac Pro.
They reference that the MacBook Pro M1 are 2.8x faster at compiling code than the MacBook Pro 13-inch Mid 2019 i7-8557U 1.7GHz. So we can scale 2xA14 to an M1, with multicore Geekbench.
Each core loses about 1-1.5% for every extra core, to the interconnect.
3 * (3826 * 2.8 / 2 / 4000) / 0.97 ~= 4.14
For MacBook Air it's probably limited to ±11W (2xA14), with the MacBook Pro maybe twice that (~33% higher clock eats quite a bit). E.g. the Air will clock down to 3GHz pretty soon. The Pro should have lots of thermal room, given that it was designed for a 28W Intel CPU.
They reference that the MacBook Pro M1 are 2.8x faster at compiling code than the MacBook Pro 13-inch Mid 2019 i7-8557U 1.7GHz. So we can scale 2xA14 to an M1, with multicore Geekbench.
Each core loses about 1-1.5% for every extra core, to the interconnect.
3 * (3826 * 2.8 / 2 / 4000) / 0.97 ~= 4.14
For MacBook Air it's probably limited to ±11W (2xA14), with the MacBook Pro maybe twice that (~33% higher clock eats quite a bit). E.g. the Air will clock down to 3GHz pretty soon. The Pro should have lots of thermal room, given that it was designed for a 28W Intel CPU.
Last edited:
It'll definitely be interesting to see how it holds up to Ryzen CPUs, since Intel isn't really the benchmark for CPU performance anymore.
Its also in-line with what they’ve been doing on the iPad and iPhone... which also simplifies things for consumers (not tech nerds) and I suspect will simplify their lineup.
GHz and turbo boost speeds are just confusing for your average punter - even more so now when clock speeds don’t tell the real story anymore and people are wondering why they should buy a new machine with a lower clock speed than something they already own.
GHz and turbo boost speeds are just confusing for your average punter - even more so now when clock speeds don’t tell the real story anymore and people are wondering why they should buy a new machine with a lower clock speed than something they already own.
Asking for MHz or GHz is obvious for comparison. Short answer, you shouldn't for now.
So I wonder if they aren’t going to clock down the M1 in the Air or just let let thermal throttle most of the time? That is kinda hard to believe they would do that
That would be completely ridiculous. You just suggested that the M1 outperforms latest AMD Zen3 desktop CPUs by 30% in single core.
CPU monkey in this as useful as an actual monkey. Probably run by monkeys as well since they decide to post this kind of unverified nonsense.