The A12 Myth and Bench-marking
I know this an Apple forum and I'm probably going to be flamed - but nevertheless :
When this chip launched it was supposed to a massive powerhouse that was decades ahead of everything and 90% faster / 90% better / 90% more efficient. Even now, there are articles littering this forum about it being two generations ahead of everything. But here's the thing when actually doing research :
* It's quick - there's no doubt about that - but it doesn't seem to be consistently quick. In real world examples, it open around half the apps tested faster - but the other half (Facebook, YouTube, NBC, Netflix, Spotify etc) is actually slower. Is there a reason ?
* Then you get people (and articles) saying how the older A11 is quicker than anything as well. Has anyone seen the A11 (iPhone X) against the Kirin 980 at all? It opens 100% of all apps tested faster. Every. Single. One
So it got me wondering how the articles can be so wrong, so often and I dug deeper. It seems that it's all being based on a theoretical benchmark test. Some people don't really understand bench-marking - but effectively this a theoretical number that can give an indication of performance. One massive thing to keep in mind - cross platform bench-marking is not accurate. One can use the same hardware with a different OS, and get different results. In desktop terms, running Linux and Windows yields different results gives different figures with the identical CPU, primarily the way things are handled on the OS has lots to do with it ...
Some manufacturers can actually cheap with benchmarks ie OnePlus where these scores can actually be manipulated - is that a true reflection of the CPU if a manufacturer can alter these theoretically numbers?
If there is some other test that shows why this A12 (and the older A11) are light years ahead of everything in existence, perhaps I missed it and welcome a post. Also, is it possible to ban a single user from this thread so that it doesn't get contaminated? I'd prefer to keep madtiger and hid comments out of this particular thread.
When this chip launched it was supposed to a massive powerhouse that was decades ahead of everything and 90% faster / 90% better / 90% more efficient. Even now, there are articles littering this forum about it being two generations ahead of everything. But here's the thing when actually doing research :
* It's quick - there's no doubt about that - but it doesn't seem to be consistently quick. In real world examples, it open around half the apps tested faster - but the other half (Facebook, YouTube, NBC, Netflix, Spotify etc) is actually slower. Is there a reason ?
* Then you get people (and articles) saying how the older A11 is quicker than anything as well. Has anyone seen the A11 (iPhone X) against the Kirin 980 at all? It opens 100% of all apps tested faster. Every. Single. One
So it got me wondering how the articles can be so wrong, so often and I dug deeper. It seems that it's all being based on a theoretical benchmark test. Some people don't really understand bench-marking - but effectively this a theoretical number that can give an indication of performance. One massive thing to keep in mind - cross platform bench-marking is not accurate. One can use the same hardware with a different OS, and get different results. In desktop terms, running Linux and Windows yields different results gives different figures with the identical CPU, primarily the way things are handled on the OS has lots to do with it ...
Some manufacturers can actually cheap with benchmarks ie OnePlus where these scores can actually be manipulated - is that a true reflection of the CPU if a manufacturer can alter these theoretically numbers?
If there is some other test that shows why this A12 (and the older A11) are light years ahead of everything in existence, perhaps I missed it and welcome a post. Also, is it possible to ban a single user from this thread so that it doesn't get contaminated? I'd prefer to keep madtiger and hid comments out of this particular thread.
Comments
That opening apps is all there is to determine how good a chip is? So, if Kirin 980 or S855 uses their big cores to open lightweight apps while Apple uses the small cores due to algorithm, then that's doom for Apple?
The actual gameplay is not important? The actual 4k video editing etc. is not important? Basically, the actual usage of the app is not important in your little head?
Sorry, but your post is just moronic and shows your total misunderstanding. BTW, thanks for posting this......good job in showing your ignorance in full glory! Nice!
(Of course you want me out of this thread because i have proven you wrong over and over again. LOL)
You're not technical and have no idea of technology - which isn't a bad thing, because it isn't your forte.
I'd just like a few of the smarter people to try help.
I mean this comment aboveb is extremely bizarre. Apple create the fastest chip in the world and then purposely slow it down with algorithms, in order to make it *not* the fastest at actual opening apps? That blows my mind that you believe that ...
Your tech knowledge revolves around app opening speed, instead of actual app performance! Stupid.
Your ignorance is bizarre indeed.
Anandtech is wrong? You know about real world testing than Anandtech?
Your ignorance is profound.
May Santa bring you a clue!