Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Quad core isn't obliteration. Even Octo core is passe.

Obliteration would be an Axx chip with enough cores to run 1000's of compute threads, like a Tesla or Pascal GPU, but inside a cool battery-sipping iPhone power envelope. Plus a couple blazing hyper-turbo single-thread processor cores on the side for legacy code. That would be Obliteration.

Which is exactly what they are doing. The Israelis are top notch in chip design.

2 huge CPU cores and massive numbers of GPUs is the name of the game (6 already in A9). The GPUs will increasingly be used for computing tasks, mark my words.
 
Best Upgrade ever? I say definitely!:

iPhone 5

Bigger screen
Better Front Camera
LTE
Lightning Connector
Better Battery
new Processor (+160% speed)

iPhone 5s

EarPods (bundle)
TouchID
Better Back Camera (True Tone Flash)
M7 Motion Processor
Better Battery
new Processor (+65% speed)

You've forgot more RAM for the iPhone 5. Also, EarPods were first bundled with the iPhone 5 as well, not the 5s.
 
Make calls, look at emails, text people and look at the internet, facebook and twitter. Yes, a great need for "desktop-class computing performance".

Ever heard of "hurry up and wait"? Having raw performance allows a processor to complete it's instructions and go back to sleep sooner. This adds up to increased battery life over a design then spends twice as long crunching (and less time sleeping).
 
  • Like
Reactions: rGiskard
It doesn't really get crushed in multithread, even with 6 less cores.
Considering AMD bested Intel in Multi core for a long time, we all know how multi core is "important" to actual real world apps on a desktop. On a mobile phone, its actually next to useless cause very few apps use all the cores efficiently.
For god's sake, the OS itself doesn't even use these cores efficiently as Safari on the Iphone creams any other browser on Android! let alone a random app.

Well said. I've always suspected that 4+ cores on a smartphone is primarily useful for marketing.

OTOH, 4 cores on an iPad Pro could be used by video editing apps, maybe even some graphics apps. At least if the power is there then some crazy developer can put it to use.
[doublepost=1473198479][/doublepost]
What are you talking about? Thats just false.

Best Android Device - 5212
iPhone 7 - 5495

https://browser.primatelabs.com/android-benchmarks

You're right, I confused it with the 6S score.

This iPhone 7 is even more impressive since Apple has traditionally been behind on multi-core performance! If the camera is as good as I hope it is then I may have an extra expense coming this winter!
 
For iPads it's time for Apple to take their SOC prowess to extremes and fiinally intro a quad core Ax SOC that obliterates the competition. We need ludicrously excess power to spark development of the next killer iPad app - an app that cannot be duplicated on Android because their hardware is impotent.


Most mobile developers build with cross platform in mind. This unlikely to happen. To cut off a market is to cut off potential revenue sources.

It also cut off sales within apple users. People on the older hardware. As some like me see no reason to upgrade. yay the new stuff runs AES really fast. this is not a task I do 24/7.

Apps I use will never come to iOS because of iOS's restrictions. Or pointless to cross over. Which is fine current ipads run apps I use just fine, have the MBP for the other stuff. Right tools for the right job mindset.

Killer app that won't run on my mini bought last year (and I think is more than fine even now)...won't be a sale. from me or other content owners of current hardware. Hard market to blow off given apple keeps up hardware support a fair amount of years.


Clearest working example is the games industry in PC's. They develop based on hardware a few years back and not bleeding edge. Games could be better if they did that in all honesty but it risk of less sales from the market base still on 2-3 year old hardware.

Make the game to have even the mighty titan humbled and titan owners will say thank you for breaking the boundaries of gaming is the same game that has non titan 2-3 year old owners going no thank you for a game we can't play. Latter is potentially a very large market to tell and go pound sand. So game publishers and devs avoid that as much as possible.
 
iPhone bench marks are never faked. Only Android and especially Samsung scores are faked.

I know you're being sarcastic, but the funny part is you're actually right. Ooops. Bet you didn't mean for that to happen.

There are a LOT of faked low iPhone 6S scores. Like the ones I linked below:

http://browser.primatelabs.com/v4/cpu/5953
http://browser.primatelabs.com/v4/cpu/183174

The first one is done better because it actually lists the correct processor. The second one is a poor fake as it shows Android as the OS and the wrong processor. Imagine that, some idiot benchmarking an old Android phone, changing the device identifier to "iPhone 6S" and uploading their result to try and trick people into thinking the iPhone 6S is really that slow. I can't describe how pathetic that is.

Now you're probably thinking "so a couple idiots did this - who cares". Well, no. You have to get to page 100 (and at 25 results per page that's 2,500 fakes) before you find results that are plausible. That's a lot of idiots on some crusade to try and lower the iPhone 6S score. I'm not sure what method they use (maybe they jailbreak an iPhone and tamper with how it runs to generate such a low score) but there's a lot of people doing this.

Apple is not the only one this happens to. The Samsung S6 and S7 (unfortunately, the S7 doesn't show up in a search for Geekbench 4 for some unknown reason) also get fake low score results. Probably from other Android users that hate Samsung. Although they are nowhere near as many as Apple. On the other end Samsung also gets ridiculously high scores posted. Even ones with Intel desktop processors that identify themselves as a Galaxy S6/S7. It's the same types of idiots trying to raise the score of their device by posting faked high scores.

It's also interesting to note there are zero (zero as in 0) faked iPhone high scores. The highest scoring iPhone 6S at Geekbench is only about 5% faster than the average score. So those are legitimate scores probably run on a device with absolutely nothing else running, all Apps closed and a cool phone that hasn't been doing anything to get warm.


The up side to this cheating? Geekbench obviously analyzes the results they get. And they don't post a device to their charts until some minimum number of tests are run (considering it takes awhile for a new device to show up this number must be large - as in several thousand). And they are obviously discarding results that are implausibly low or high. The end result is the score posted for a device is a fairly accurate average of numerous devices. Which makes me laugh at all the people uploading fake results, because it's not doing anything to affect the scores. I guess none of them never took Stats 101.

Edited for clarity/grammar.
 
Last edited:
  • Like
Reactions: ErikGrim
Dangerously approaching my 2011 i7 MBP.
Not really. Geekbench scores are not relevant.

Check out FLOPS numbers:
iPhone 6s.GPU 170 GFLOPS (about the same as a mid range AMD GPU from 2006; a PS4 has 1800 and it was underpowered even when it was launched a few years ago)

iPhone 6s.CPU: 1.2 GFLOPS (the i7 in your laptop should have around 50). :)

There's a reason why an ARM SOCs cost a few bucks and intel cpus or amd/nvidia gpus cost hundreds.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.