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.
Controlled Testing Supports Apple's Claim of Nearly Equal iPhone 6s Battery Life With TSMC and Samsung Chips
- Thread starter MacRumors
- Start date
- Sort by reaction score
The article claims that Samsung chip lags behind TSMC under 60% CPU load, with the maximum CPU load being 200%. To me, 60% isn't an unrealistic load to put on the CPU, I'm sure you don't have to go out of your way do that if you are a gamer or use phone to edit. While I wouldn't use those apps for 5 hours straight, for the duration that I am using, the Samsung chip will be draining more.
As for their sample of two phones, it's consistent with the thousands of phone tests on geekbench's website, and thus far out of thousands, they all show Samsung significantly behind TSMC on higher CPU load.
So what apps can cause high CPU load (60% total) and would you use them?
And for those who don't plan on replacing in 1 year, iOS 10 will most likely be putting more CPU stress on iphone 6s as they always do for older generation phones.
As for their sample of two phones, it's consistent with the thousands of phone tests on geekbench's website, and thus far out of thousands, they all show Samsung significantly behind TSMC on higher CPU load.
So what apps can cause high CPU load (60% total) and would you use them?
And for those who don't plan on replacing in 1 year, iOS 10 will most likely be putting more CPU stress on iphone 6s as they always do for older generation phones.
There are several issues about the so called 'difference'.
First, electronic components varies from chip to chip. Even the battery capacity varies from one battery to another. Typically the difference of battery capacity is even larger than 3%. All the so called 'tests' people did so far had proven nothing at all, because the sample base is too small. You'll have to do strictly controlled tests on thousands of devices for a meaningful statistic number.
Second, the CPU is not the most significant power consumer in a mobile device. Screen and wireless module draws even more power comparing to CPU. Thus it makes the difference of CPU irrelevant for 'NORMAL' usage, such as gaming. If some game makes a powerful CPU such as A9 running at its peak, I can hardly imagine how it would perform on older devices like iPhone 5, which the CPU score is only 1/10 of A9.
I'd say the WiFi browsing test is the one that closest to normal usage. But no one can guarantee that your phone would be 3% superior or inferior comparing to your friend's phone. Chances are that you'll get a TSMC A9 phone whose battery life is even worse than a Sammy A9 phone.
When we're talking about silicon, 'random' plays a significant role everywhere.
Seems I lucked out with a Samsung in my 6s+.
I have no doubt that the Samsung CPU uses more power based on tests thus far. It won't take long for more YouTube videos demonstrating software side by side such as games lasting longer on the TSMC.
Regular usage as implied thus far means the average. Is everybody mr average? Unlikely, some will be much heavier users, some much less, and some right in the middle fitting the profile. Majority will not be effected, but a good deal will.
In conclusion, this phone is going back as I fit the heavy user profile and I have the right to demand the best for my buck.
I have no doubt that the Samsung CPU uses more power based on tests thus far. It won't take long for more YouTube videos demonstrating software side by side such as games lasting longer on the TSMC.
Regular usage as implied thus far means the average. Is everybody mr average? Unlikely, some will be much heavier users, some much less, and some right in the middle fitting the profile. Majority will not be effected, but a good deal will.
In conclusion, this phone is going back as I fit the heavy user profile and I have the right to demand the best for my buck.
I dont play game on my Ip6s+ .. as a heavy web surfing user, I noticed my TSMS phone runs much cooler temp. TSMC battery life is phenomenon compare to my previous Sammy. I agree it wont be long before youtube videos will prove the marginal difference in battery performance
It may result in a recall if it indeed proves to be an indisputable issue. Not sure Apple can send rubber bumpers this time, perhaps a battery extension case? Lol.
I don't think so... this is not a huge issue and people are just making it an issue.
Just waiting to see when I can order my shinny new AppleTV. Been a boring week so far for any real rumors.
The problem with your singular example is that no games peg the CPU at 100%. GPU - sure, but if you notice, GFXBench tests end up the same.
^ Stereotypical apologist & fanboy answer.
Sure, I shouldn't panic about having my 1 year old iPhone 6 Plus slowed down by iOS9? I should just accept something like that, you're right.
Maybe you are right, I should look somewhere else. Somewhere where if I complain about a problem I have with my complaint heard and discredited by pompous people
^ Stereotypical apologist & fanboy answer.
Sure, I shouldn't panic about having my 1 year old iPhone 6 Plus slowed down by iOS9? I should just accept something like that, you're right.
Maybe you are right, I should look somewhere else. Somewhere where if I complain about a problem I have with my complaint heard and discredited by pompous people
If you have a problem with your iPhone, doesn't mean everyone has.
There are millions of iPhones. And most of them working fine. Wake me up when there is a real issue for the majority of devices.
Until then, stop crying for momma.
And calling me a fanboy for no reason (that doesn't get old apparently) got you a nice place in my ignore list.
Sorry everyone, I still don't want a Samsung chip.
Yeah I find it redundant for that reason.
His only solution now is to either get the iPhone 6S plus or no phone at all. People are assuming that a TSMC cpu solves all battery problems. If your usage doesn't change don't expect miracles to happen.
If battery life is critical you're better off buying a battery case like I did. I only plug in once a day and it adds protection. Plus the phone doesn't feel like it's going to slip out of my hands.
Unless TSMC can match production to the huge demand I don't see this happening. Otherwise you're going to have shortages for months. And no telling what type of cpu tech will be in the iPhone 7.
No, they will do what they should have done and split the chip shipments between the 4.7" and 5.5" - then battery tests would not have been comparable at all.
It was already established that the battery differences are on CPU side and not related to GPU. So testing GPU will not get you any result lmao. For god's sake have a look at single geekbench score. That's the only test which is CPU bound and difference is clear.
It is ok if you have a Samsung chip yourself but most people here know that CPU and GPU are two separate buildings in the same housing society.
Totally agree, this Ars Technica article just shows the GPU parts are not much different, but all 3 tests don't really use the CPU side of the chip! There is no way for me to blieve Ars Technica didn't receive a super big cheque to write this.
Yes. The rabid fanboism just ramps up to alarming levels when something like this happens every year. Last year it was bend-gate (which was real) but most pepole straight dismissed it till their own phones bent.
I hate to be a cynic but here cpu differences are extremely clear. Yet some people will harp on it saying it's a non issue for them. If I have to pay $1256 in my country just to get a 64GB variant of 6S Plus, I would want perfection.
Is Apple giving me a discount for giving me an underperforming chip? Big fat NO. So naturally I will keep returning every phone till I get a TSMC chip. It's not our fault if Apple did a half arsed research about CPU sourcing.
at least you can RETURN your phone where you live for that reason.
in Israel they'll laugh at my face.
What you are reading in this article, the result of benchmark shows, CPU is only reason of batter issues not GPU even in same APU