Kangal, 14 Oct 2023Remember, you're the one slinging insults here. I merely called you out, because you'... moreI just called you as is when you made a fitting statement. Idc about the other stuff,but sure go on.
Mediatek sux, 11 Oct 2023Did you finally lose it?
Im only talking about this statement of yours "So if they throt... moreRemember, you're the one slinging insults here. I merely called you out, because you're clearly biased. Theoretical expectations aren't always reliable, we need to wait for the software update that Apple has planned.
My point has been consistent throughout, the A17 is barely an upgrade over the A16, based on early CPU testing.... so there is a lot of cause for concern if performance gets limited, we may see the first time Apple introduced a new A-Silicon that is NOT a straight upgrade (moreso sidegrade). Except the iGPU, they've made big improvements there and potentially are ahead of the QC 8g2 depending on the game, application, and it's optimisation. So credit must be given there.
Kangal, 10 Oct 2023There, you just proved my point for me!
If you KNOW that the performance curve is not 1:1 you... moreDid you finally lose it?
Im only talking about this statement of yours "So if they throttle it, it won't be as fast or as efficient as the A16 Bionic" where you just made it like you were sure of it. But if you go look at the curves by geekerwan you'd know the a17 is clearly higher performing at the same power usage atleast in multicore. If you try to draw a curve in the other graphs, you'd atleast match the a16 in gpu and single core.
Its already more efficient in genshin test done by geekerwan.
Mediatek sux, 10 Oct 2023Telling to take a look at performance efficiency curve was my first comment, which is what you... moreThere, you just proved my point for me!
If you KNOW that the performance curve is not 1:1 you would then understand that limiting performance to prevent throttling might mean reducing performance higher than normal. So much for calling me dumb, when you couldn't do the simple maths in front of you.
If the A17 had a +30% performance lead to the A16, then sure people would NOT be worried. However, it has slim margins with only 6%-9% lead. You telling people to calm down is not effective, and shows bias. People should be angry and they should be worried. Companies are NOT your friend, especially Apple.
PS: we can be optimistic and see perhaps a big software update, (drivers, kernel, system), might be able to improve the base performance of the A17 devices. But something as mundane as energy limit won't really be improving things. Isn't it peculiar that I can distill these concepts down, but you haven't made any addition or ideas to the discourse..... funny that !
Kangal, 10 Oct 2023No, it's not.
But is it so hard for you to understand that limiting the performance may ... moreTelling to take a look at performance efficiency curve was my first comment, which is what you originally replied to.
Idk how you missed that. And if you went and looked at it you'd know you're wrong.
Mediatek sux, 10 Oct 2023Is it that hard for you to admit your comment of "So if they throttle it, it won't b... moreNo, it's not.
But is it so hard for you to understand that limiting the performance may prevent throttling, but it might not necessarily aid efficiency. Both things can be right, you do not see that? You need to understand that this all lies on a performance-curve, so things are not 1:1 as you expect.
...so you still want to preach to everyone, and call me dumb and ignorant?
(i mean it's your voice, go bananas)
Kangal, 10 Oct 2023You cannot just factor out the lithography, it doesn't work that way. The end-user resul... moreIs it that hard for you to admit your comment of "So if they throttle it, it won't be as fast or as efficient as the A16 Bionic" is wrong/ignorant/dumb?
Mediatek sux, 09 Oct 2023Again, im not commenting on tsmc node improvements, but your dumb/ignorant comment. A17 in sin... moreYou cannot just factor out the lithography, it doesn't work that way.
The end-user result is the culmination of software, system, architecture, and lithography.
You keep saying my "dumb" comment but you never elaborate on it. Why attack me, instead of my opinion? Are you not able to protect your opinions against mine?
You seem to forget, and I keep reminding you, the performance difference between the A17 and the A16 is between 6%-9%. That is not really an improvement, it's almost in the margin of error. Still. If you had that type of difference and you decided to make the difference even smaller, what do you get? Please think and tell me. Hint Hint, I have already given you the answer in a previous comment.
You seem to be one of the few people who actually are happy with the iPhone 15 Pro's efficiency and performance. Read the room.
Kangal, 09 Oct 2023Well, if you looked at the graph it's not much of an improvement. Right now the A17 is o... moreAgain, im not commenting on tsmc node improvements, but your dumb/ignorant comment. A17 in single core will atleast match a16 even if downclocked, its noticably more efficient in multicore due to the improved little cores, the gpu is slighlty more efficient.
Mediatek sux, 08 Oct 2023Im not talking about tsmc not getting upto expectations, but your comment of "So if they ... moreWell, if you looked at the graph it's not much of an improvement.
Right now the A17 is only 6%-9% faster than the A16. And it's doing that by using a bit more power and throttling earlier. So you understand where I'm coming from ?
If Apple holds back performance too much, it will just be equivalent to the A16. So it will be a sidegrade and not an upgrade. Making the whole ordeal very disappointing. Remember this is a full-node jump. The last times we had full-node jumps we had anywhere from "good to massive" improvements, such as 45nm to 28nm, or 28nm to 14nm, and 14nm to 7nm. Whereas with the middle or half-node jumps we saw anywhere from "none to good" improvements, such as 28nm to 20nm, or 14nm to 10nm, or 7nm to 5nm.
The figures I quoted in my previous comment come directly from TSMC, and you can see this is a full-node jump and it is pretty good, not the best, but definitely a decent upgrade. We just are not seeing any if those promises delivered, at least not this year. So perhaps Android with QC 8g3 will surpass iPhones with the A17. Or Qualcomm will also be affected, meaning the QC 8g3 will be a sidegrade of the QC 8g2, and Android will still languish behind iPhones. However, there is a pretty high certainty that the A18 and the QC 8g4 will at least be significant upgrades over the A16 and QC 8g2 respectively. You can't say the same thing about the QC 8g1 versus the QSD 865, or the A17 versus the A15. Those theoretical upgrades do not always translate to real-world improvements.
Kangal, 08 Oct 2023Maybe instead of calling people dumb, you can do more research.
And that Geekerwan video, at ... moreIm not talking about tsmc not getting upto expectations, but your comment of "So if they throttle it, it won't be as fast or as efficient as the A16 Bionic"
If you look at the spec graphs and know how power draw isnt linear to perf you'd know the a17 is gonna be slightly higher performing if its throttled to the maximum power usage point of the a16. In multicore it is obviously more efficient.
Mediatek sux, 03 Oct 2023Actually watch the graphs and plot lines and you'll see. He even does it for you during t... moreMaybe instead of calling people dumb, you can do more research.
And that Geekerwan video, at the timestamps you posted, listen closely to what he says!
" the power efficiency isn't that bad at all. Comparing to A16 at same performance, the A17 actually sips slightly less juice. But if you were hoping for a drastic improvement in efficiency, well, it is not here yet. Seems like we'll have to wait longer. "
.... now, take in the context of what TSMC initially promised with the 3nm node:
+15% more performance AND -30% less energy AND x1.7 more density
or at the minimum:
+10% more performance AND -25% less energy AND x1.5 more density
and at the maximum:
+18% more performance AND -34% less energy AND x2.0 more density
So we were expecting "drastic improvement" based on the modest figures above from a full-node jump. We did not see these promises delivered by the A17 chipset. Not sure it is Apple's fault or TSMC or both. But it seems likely that TSMC and Apple needs that extra year to improve the quality of their yields and chips, so that the processors can match these expectations. Right now they aren't, and Apple seems to be pumping them up with higher voltages to maintain stability, but it comes at the big problem of less performance, higher energy draw, and quicker throttling. I don't think a software update can fix this, not without sacrificing something.
DaFink, 05 Oct 2023well, everything other than gaming obviously. You get they do other stuff right. Oh who am I k... more Oops, nearly forgot I owned a Nokia 1. It does everything except gaming obviously. Oh who am I kidding, no that probably hasn’t occurred to you at all
Cpt.Power, 06 Oct 2023Well no my friend.
I love fun and humour in any kind and form.
But ultimately its your cho... moreGood for you, I have had various Moto’s in the past, mostly prior to the Lenovo acquisition, since then, they are to me the very epitome of bland, having lost anything that made the brand special in that acquisition. They are now just another of those Chinese brands that seem to love saturating the market with never ending variants and off shot devices, all barely indistinguishable from each other. And that’s just not my thing.
Again, that’s just me though. I don’t ask anyone to share my opinion, nor suggest they made a mistake in purchasing such a device. And I’m certainly not an Apple salesman, so don’t really give a stuff who buys or doesn’t buy from them.
It continues to puzzle me mind that those that neither use nor like Apple stuff tend to be the most insistently vocal about Apple on this site, it’s a weird trend that has never made any sense to me 🤷
DaFink, 06 Oct 2023I see sarcasm still isn’t one of your strong suits. And yes I am enjoying my phone thanks, des... moreWell no my friend.
I love fun and humour in any kind and form.
But ultimately its your choice and pick what you prefer and like.
For me anything which takes 2 hours to charge its battery is a no for me no matter the brand.
My current phone Motorola edge 40 pro charger from 20% to 100% in like 15 minutes and i wouldnt go to anything slower no matter what.
Cpt.Power, 05 Oct 2023Well you funny. Motorla making like 20 devices a year not 100+ in 4 months but allright. Wel... moreI see sarcasm still isn’t one of your strong suits. And yes I am enjoying my phone thanks, despite the experience most people on this site (the vast majority of which don’t use iPhones) tell me I should be having 🤷
DaFink, 03 Oct 2023Well our difference there then is I will not be handing out tips and advice because whilst I k... moreWell you funny. Motorla making like 20 devices a year not 100+ in 4 months but allright.
Well enjoy your phone then.
Game Legion, 05 Oct 2023What do you have an iPhone for?well, everything other than gaming obviously. You get they do other stuff right. Oh who am I kidding, no that probably hasn’t occurred to you at all