So Apple even with a new node only managed a 20% improvement. So a new node and clocks that means no IPC improvements. Going by shrimps leak.
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.
Leak of 3nm A17 chip - single core 60% faster?!
- Thread starter donawalt
- Start date
- Sort by reaction score
My realistic take on this, is that going from to 5nm to 3nm would enable the 60% improvement described by the leak, but Apple will ship A17 with ~20% performance boost and efficiency improvement, also because 3nm is here to stay for several years and they have to gauge the improvements for A18-A19 as well.
So an A17 benchmarking with 60% boost over 5nm would be fake news, but holding some truth nonetheless.
Of course and we don't know what Apple is cooking with the new 3nm chips. the 4.5Ghz was just a random number, what I'm saying is in single threaded workloads you could have a P-core run at something like 4Ghz - 4.5Ghz while the whole system would run at like 3.3Ghz when multi-threaded. It what Intel has been doing with its turbo-boost technology since over a decade.
BTW. Here's my Geekbench 6 benchmark of the A15. Seems quite in line with that leak when it comes to the difference between single and multithreaded performance. I'm talking about the 3000/7800 score that is.
iPhone 14 Plus - Geekbench
Benchmark results for an iPhone 14 Plus with an Apple A15 Bionic processor.
browser.geekbench.com
Last edited:
Guys, I got a new exclusive leak! Directly from the GB6 browser!
20% is the sane expectation for a new node and P-core design. It’s the uplift we saw with A14. Not to mention gains from new nodes are decreasing.
The few people trying to argue 60% is plausible are just silly or ignorant.
Last edited:
That is a good point, but despite the market share I haven't really seen many "console quality" games on mobiles even though recent devices probably are approaching the power of a PS5 (or will be soon in a year or two). Maybe they are out there, but as far as I can tell variations of Candy Crush clones have dominated for the past decade with Genshin Impact being the one exception I can think of to go mainstream. In fact, the future for AAA games seems to be cloud gaming on phones which doesn't have much to do with the hardware.
Not that I'm saying more power isn't a great thing, just that I agree with the other guy about battery life and other improvements as being more important at this point.
Even if the numbers are true, are they coming from a phone prototype or a test platform plugged into the wall? Apple always downshifts the processor in iPhones to extend battery life, so an iPhone 15 would not have the same performance.
Have you considered how Intel is able to reach those clocks, such as examining the pipeline structure vs. width? Or what the power consumption is under Turbo?
Again, this all comes down to fundamentals which you clearly don't understand.
I think it's more likely that when they turn out to be not accurate, AAPL stock dips over some "not meeting expectations" BS
They have a little more work to do in terms of
readInstruction->interpertInstruction->getResources->actuallyDoInstruction->commitResult->retireInstruction
which makes the longer pipeline necessary to run at higher speeds.
The ARM core grabs 8 words at time and partially resolves them in one cycle, handles resource allocation in the second cycle and has them in their execution queues by the beginning of the third cycle. Full interpretation occurs as part of the work cycle, within the execution unit. Commit/retirement can be complicated, but, on average, there is far less work happening at the top end.
Compilers for x86 lean toward minimizing the top-end work, but there are practical limits. ARM-type compilers optimize in different ways. But the wide dispatch scheme has a lot less to gain from cranking up the clock.
I play CoD mobile and graphics are about PS4 quality with maximum settings on a recent iPhone / iPad, CoD is huge on mobile with the lion's share of users on iOS, it might have more users than the console version.
Edit: I just found this report from last year:
COD Mobile matches user numbers with PC and console Call of Duty titles, Activision reports
COD Mobile was first released in October 2019, and since then, the first mobile title of the Call of Duty franchise has only grown in size and its close-knit community.
www.sportskeeda.com
CoD is larger on mobile than PC & console, and perhaps on of the reasons behind Microsoft interest on buying Activision.
Last edited:
Of course and we don't know what Apple is cooking with the new 3nm chips. the 4.5Ghz was just a random number, what I'm saying is in single threaded workloads you could have a P-core run at something like 4Ghz - 4.5Ghz while the whole system would run at like 3.3Ghz when multi-threaded. It what Intel has been doing with its turbo-boost technology since over a decade.
BTW. Here's my Geekbench 6 benchmark of the A15. Seems quite in line with that leak concerning the difference from single to multithreaded performance.
iPhone 14 Plus - Geekbench
Benchmark results for an iPhone 14 Plus with an Apple A15 Bionic processor.
browser.geekbench.com
I'm sure you'll enlighten us all about those fundamentals which I don't understand....
...I thought so!
Being a RISC architecture you surely understand the pipeline is much shorter than something like an x86 and thus much more flexible to a wider clock frequency range?
Last edited:
Wow, only three words in and I already knew this was fake as soon as I saw "Max Tech." 😅
That is pretty cool! I did not realize COD mobile had caught on so much. I still think there is a real lack of console quality anything on mobile that better hardware isn't going to change, but obviously more gamers out there than I thought could benefit from more power.I play CoD mobile and graphics are about PS4 quality with maximum settings on a recent iPhone / iPad, CoD is huge on mobile with the lion's share of users on iOS, it might have more users than the console version.
Edit: I just found this report from last year:
COD Mobile matches user numbers with PC and console Call of Duty titles, Activision reports
COD Mobile was first released in October 2019, and since then, the first mobile title of the Call of Duty franchise has only grown in size and its close-knit community.
CoD is larger on mobile than PC & console, and perhaps on of the reasons behind Microsoft interest on buying Activision.
If it hasn’t been made clear yet, I’ll type it in all caps for you…
IT’S NOT ABOUT BOOSTING PERFORMANCE TO THE MAX, IT’S ABOUT HAVING THE BEST PERFORMANCE PER W.
Even if hypothetically Apple could make the 15pro 60% faster, they wouldn’t.
They’d make it 20% faster… while consuming less power.
Who needs more speed? People who use their computers for actually work, rather than for posting dumb comments on forums.
I don't need battery life on my work MacBook M2 Pro. It's almost always plugged in to power. But I need all the speed I can get. On my personal MacBook Air, I want great battery life, am less concerned about speed and I'm perfectly happy with M1.
The only people who might hold off are those who always want the latest and greatest devices (which is a vocal, yet admittedly small subset of the overall Mac user base). For the vast majority of Apple users upgrading from older hardware, it really wouldn't matter whether there is an M2 or M3 under the hood, just whether the machine will work in general.
The OP and thread discusses the upcoming iPhone SoC, A17, not an M-processor. I know the forum is Apple Silicon Macs, but there you are. My comment is about iPhones.
The "power" users only comprise a small subset of the Mac user base as a whole. For the rest, battery life may be a primary concern if they are buying a laptop. Add in the fact that Apple Silicon doesn't drop into a throttled mode when off AC power (unlike Intel and AMD systems), and you can actually get both performance and battery life instead of having to choose between the two options.
No, not "Fake until proven true." This is a rumors site, and this remains a rumor until proven true. Calling it fake is some fool declaring that a rumor is untrue without any data one way or the other.
Wrong. Our default position on anything we hear from hearsay should be to believe that it is unverified, not to always default to think that it is either false or true.
I play CoD mobile and graphics are about PS4 quality with maximum settings on a recent iPhone / iPad, CoD is huge on mobile with the lion's share of users on iOS, it might have more users than the console version.
Edit: I just found this report from last year:
COD Mobile matches user numbers with PC and console Call of Duty titles, Activision reports
COD Mobile was first released in October 2019, and since then, the first mobile title of the Call of Duty franchise has only grown in size and its close-knit community.
CoD is larger on mobile than PC & console, and perhaps one of the reasons behind Microsoft interest on buying Activision.
That is really interesting, and I suspect the Mobile version is as profitable or more so than the other platforms. After all, if its a in-app purchase model like most mobile games these days, it is generating a continuous revenue stream from all those players, rather than just a one-off purchase.