Controlled Testing Supports Apple's Claim of Nearly Equal iPhone 6s Battery Life With TSMC and Samsung Chips

[Keirasplace]

So it's no big deal except you use your phone a lot for cpu/gpu intensive tasks like for example gaming. Who plays games on his phone anyways?

This GPU is way massively overpowered for 100% of games on IOS; yes, 100%.
A few dev were close to using the A8'S power, for this one, not even a mile away.

I doubt playing all those games would make much of difference unless your game of choice is asphalt; even then, I'd expect a difference of maybe 5-10% if you play it continuously till you kill the battery (which is not a typical case).

As for the CPU, since the CPU is 40% is faster than last year, and even then, few games lets face it, were close to maxing the CPU (because devs want their games to work on slower phones), all this talk about a game "maxing" the CPU is pure BS for most games. You're very very far from close.

So GPU and CPU far from hitting the max on existing game. That's it. Hope this forum stops being such a drama queen

 

[asleep]

But is the heavy benchmark indicative of normal everyday usage, probably not.

GB3 isn't taxing the CPU nearly as much as I suspected:

...the Geekbench 3 battery life test generally keeps the CPU pegged between 55 and 60 percent load for the entire time it runs, with occasional dips below 55 percent and peaks above 60 percent.​

It appears the more processor focused tasks show the highest Δ between the two... processors, aha! So if you want to mask the differences on battery drain, turn up brightness, add GPU tasks, etc.

We would remind readers that TSMC has made a very big deal about their power consumption/leakage advantage over other competitive manufacturing processes. TSMC has claimed up to a 20% power advantage in their 16nm process...​

ArsTechnica pegged that difference at 28% using GB3, which is on the high variance side of TSMC's prior stated advantage.

The good news is, it appears we won't face this disparity in the next generation...

It's no surprise that Apple is going all TSMC with the A10​

 

[Thunderhawks]

It does not matter how many tests are done, some people will still lose sleep over this.

I bet somebody could make a lot of money selling a chip kit with 1 TMSC and 1 SAMSUNG chip.

So, every user who feels there is something wrong can switch their phone to the chip they think let's them sleep at night.

Or, start swapping out chips in the malls with screen repair booths/stands.

All the hypochondriac complainers would probably go insane, when they cannot decide which one to use/put into their phones.

That then would confirm the non issue we have here.

Plan B is for SAMSUNG to buy TSMC and make one combo chip at 15nm (even if it doesn't exist) manufactured by both factories.

 

[Gregg Thurman]

Yea, technically correct reporting of processor deltas.

Its a fact of life that processors, produced by the same firm and cut from the same wafer, will perform differently than its wafer neighbor. That's why performance specs are given for the lowest common denominator (performer) and production specs are for a range of performance results. The tighter the range the higher the cost per processor and vice versa. The deltas given between the TSMC and Samsung processors are technically minuscule, to the extent that they will not be noticeable to the overwhelming majority of users, and barely noticeable (if at all) to a very, very small minority of users.

I can't help but think those looking for an exploitable controversy re: iPhone are stretching the bounds of credibility. This "controversy" is a non-issue.

 

[zone23]

Nothing to see here. My wife has a TSMC 6S and she just mentions that it doesn't last as long as the 6 did.

Thats not the question the question is will a 6s TSMC outlast a 6s Samsung if you were say paying a game none stop until the battery died and if so how much longer did the one last than the other.

 

[kironin]

But will people believe it? Of course not.

Of course! LOL Screw Apple, I still want the TSMC phone! it also runs cooler according other testers.

 

[Thunderhawks]

This GPU is way massively overpowered for 100% of games on IOS; yes, 100%.
A few dev were close to using the A8'S power, for this one, not even a mile away.

I doubt playing all those games would make much of difference unless your game of choice is asphalt; even then, I'd expect a difference of maybe 5-10% if you play it continuously till you kill the battery (which is not a typical case).

As for the CPU, since the CPU is 40% is faster than last year, and even then, few games lets face it, were close to maxing the CPU (because devs want their games to work on slower phones), all this talk about a game "maxing" the CPU is pure BS for most games. You're very very far from close.

So GPU and CPU far from hitting the max on existing game. That's it. Hope this forum stops being such a drama queen

Nooooooooo we need to die a little every day with insignificant Apple problems. If worse comes to worse we'll make some up.
The other day I got an upset stomach eating a candy bar while on my iPhone. Coincidence?

Went to the genius bar and they asked me if I had taken off the wrapper. Problem solved.

 

[thegrants82]

QUOTE="MacRumors, post: 22071478, member: 3"]


Last week Apple addressed concerns over battery life discrepancies between the TSMC and Samsung A9 chips used in the iPhone 6s and the 6s Plus, suggesting a real world battery life difference of only two to three percent between the chips despite Geekbench battery tests that had pointed towards a bigger gap.

Ars Technica has now conducted controlled battery life tests on two iPhone 6s models, one with a TSMC chip and one with a Samsung chip, and its findings line up with Apple's statement. The site used two AT&T models, both with SIM cards removed and screen brightness set at the exact same level.

In Wi-Fi browsing, WebGL, and GFXBench tests, there were performance differences mostly in favor of the TSMC iPhone, but the variation between the two phones was slight. There was a more significant performance difference on the Geekbench 3 test, but as has been covered earlier, that test is not reflective of real world usage.

Aside from the Geekbench test, which saw a battery life difference of 28 percent between the two devices, the TSMC iPhone and the Samsung iPhone scored within two to three percent of each other. In Ars' opinion, in the real world, there's going to be little difference between a Samsung iPhone and a TSMC iPhone.According to Apple, the comparison of just two devices is too small of a sample size to make meaningful predictions about the performance between iPhones with Samsung chips and those with TSMC chips. As Ars points out, no one but Apple has the resources to collect enough data from enough devices to get an accurate picture of performance.

Ars Technica's findings do line up with what Apple says it has seen in both lab testing and data gathered from hundreds of thousands of customers. Apple has said the variation between the TSMC and Samsung chips it has observed is "well within its manufacturing tolerances" and is a level of variation that can be seen between devices that have the same chip.

Article Link: Controlled Testing Supports Apple's Claim of Nearly Equal iPhone 6s Battery Life With TSMC and Samsung Chips[/QUOTE]
My 6S has the TSMC but if I had the Samsung chip I would return it.

 

[Max(IT)]

so basically this controlled test said nothing.
Difference only in Geekbench and just one sample tested (so statistically not relevant) ...
The point is, if we want a credible base of data we have to believe in Apple.

 

[Adithya007]

Why do people need to worry if they've got the TSMC A9 or the Samsung A9 chip? I'm sure that both of them will be able to add filters to the Foodie Pic and upload it to Instagram!!!

 

[Thunderhawks]

Exactly. people are idiots.

And on top of that they buy Rose Gold phones!

 

[SSD-GUY]

It's not a big deal.

The differences are subtle if at all. A Year from now, the 6s will be the discount phone, and the battery and performance will be ousted by the 7. So, details like this are really are en passe and not worth your time. If I got a phone with like, a cosmetic defect, I would absolutely return it, but for things like this, no. Esp with the upgrade programs now in place, I don't fret as much about these things. It is my favorite thing, ever. I will be a subscriber on the Apple upgrade program for life. It was made for me.

It is a big deal. Why do I need to wait for another iteration of the iPhone to surpass my 6S for it to stop being an issue?

One chip performs better than the other. Under load or not isn't the question, what it boils down to is that one chip is more efficient than the other.

 

[Sgt.ThroatPunch]

So it's no big deal except you use your phone a lot for cpu/gpu intensive tasks like for example gaming. Who plays games on his phone anyways?

Sexist pig.

 

[MikeKlump]

Why do people need to worry if they've got the TSMC A9 or the Samsung A9 chip? I'm sure that both of them will be able to add filters to the Foodie Pic and upload it to Instagram!!!

I don't think the people upset are casual users/consumers. In comparison to last year's bendgate this "controversy" is very minor. It's a lot easier to talk about phones bending on the evening news, you lose a lot of people's interest when you start talking about chips/processors.

 

[yep-sure]

Read MacRumours forum on 05/10 - most posts saying "OMG I WANT TO RETURN MY TSMC CHIP FOR A SAMSUNG"
Read MacRumours forum on 11/10 - most posts saying "OMG I WANT TO RETURN MY SAMSUNG CHIP FOR A TSMC"

I'm pretty sure this isn't the first phone that Apple has used multiple chip suppliers for, right? Either way, I'm getting a 6S tomorrow, and have no idea what chip I'll end up with.

 

[Sgt.ThroatPunch]

I don't think the people upset are casual users/consumers. In comparison to last year's bendgate this "controversy" is very minor. It's a lot easier to talk about phones bending on the evening news, you lose a lot of people's interest when you start talking about chips/processors.

The chip might not interest some people, but battery life affects everyone. I'm not saying it is huge difference for the average user, but there is a difference.

 

[Max(IT)]

The heavier Geekbench test, on the other hand, showed the TSMC phone lasting an average of 28 percent longer than the Samsung phone.​

Are you spending your day running Geekbench ? Or are you actually using your iPhone ?

The sample size of only testing two phones is still ridiculous from a statistical point of view.
e.g. what is the standard deviation of the variations from phone to phone with the same hardware?

Too much common sense, mate...

I'm saying that since the beginning, but the forum like the hysteria much more.
A "good Samsung" could be better than a "poor TSMC", even if a real difference between the twos exist.
And I can't believe Ars Technica, one of my favorite web sites, didn't put that as an enormous disclaimer first.

It's hard to say what real world CPU usage will be. If I spend a lot of my day playing while(1).app, this could be a big deal.

Who says your game is something like Geekbench ?

 

[Sgt.ThroatPunch]

I have two 6s iPhones, a Sammy and a TSMC. I'd like to do my own testing and record the results. Any tips?

 

[Asclepio]

TSMC's phones are over 1,500 bucks on ebay

 

[mr.bee]

I bet Samsung is behind this chip gate! Their chip won't get the blame, it's Apple who decides to work with 2 different chips!

 

[MikeKlump]

The chip might not interest some people, but battery life affects everyone. I'm not saying it is huge difference for the average user, but there is a difference.

Agreed. I just haven't heard anyone outside of my techie friends talking about it.

 

[Sgt.ThroatPunch]

Agreed. I just haven't heard anyone outside of my techie friends talking about it.

Me either. It's just the trend. Many of us spend so much time discussing the release of the phone, followed by the obsession with tracking information and shipping concerns. Suddenly the phone arrives, and there is a void of things to talk about.

 

[Max(IT)]

I'm not even going to check which A9 I have. I just don't want to know. I'm not exchanging my iPhone based on whether it has a Samsung or TSMC chip so what's the point? Am I right?

You are right, but this is MacRumors, and people have to open a #gate about something every year ...

Has someone actually done a test that isn't running a benchmark but instead actually running a processor intensive app/game? To me running a geekbench test isn't real world usage.

This.
I'd like to see the difference running a REAL game, not a synthetic benchmark claiming to be similar to a game.

 

[roeiz]

I have two 6s iPhones, a Sammy and a TSMC. I'd like to do my own testing and record the results. Any tips?

I'd like to see the difference running a REAL game, not a synthetic benchmark claiming to be similar to a game.
try a rather demanding game.. many thanks!!

 

[Sgt.ThroatPunch]

TSMC's phones are over 1,500 bucks on ebay

By "phones" I assume you mean the one idiot who posted his/her phone?