My real world experience with TSMC vs Samsung

[HEK]

So your "real world experience" consisted of running the same benchmark test that everybody else has already ran, and yielded the same result everybody else already knows. THE TSMC CHIP LASTS TWO HOURS LONGER THAN THE SAMSUNG CHIP ON THE GEEKBENCH BATTERY TEST!!! Just in case there are a couple people left in the world that haven't heard this yet.

The question is: How long does the TSMC chip outlast the Samsung chip in REAL WORLD use? As in, day-to-day regular usage. This is harder to test, as it requires one of each phone, doing the exact same things, at the exact same time. This also depends on the user's primary usage. For example, I don't shoot 4K video, then edit it in iMovie. So, that kind of test won't accurately determine which chip is better for my usage. How much of a difference is present, if any, during extended Safari browsing?

Yes, I have the Samsung chip in my 6S+. My wife has the TSMC chip in her 6S. We both have been happy with our phones. Coming from a 6, I have no complaints with the battery life of my 6S+, and I have no plans of returning it.

Exactly!
The test is flawed because the smaller Samsung chip running continuously heats up more and consequently uses (wastes) more power. The larger TSMC chip which runs a very slight bit slower and is larger surface area provides slightly more efficient cooling and does not get as warm. This means less power waste from tunneling.

The test is extremely artificial subjecting the chip to extreme loads it does not see in everyday use of any kind. Chips are normally seeing bursts of use with resting cycles. This means that the heat build up seen in the test and consequent increased energy consumption is not experienced outside of the test.

A truly relevant test would be to look at several thousands of phones under various conditions, locations, screen brightness, etc. tabulate a statistical mean and standard deviation and produce a six sigma bell curve of total battery durations. Do this for each chip phone calls mbination and compare those bell curves. Now you would have a meaningful test result to make the determination of which chip is superior in daily use.

Without this kind of data, the tests on your part with individual phones are completely irrelevant and lead to totally false assumptions as they have for the OP and some others.

The only party in a position to gather such data is Apple itself via the user experience data download it does from the each phone daily. Unless you have turned this feature off. This is how Apple arrives at the 2-3% difference it has stated publicly.

Since this is also within the statistical variation between the same chips made from silicone wafers. There is NO significance to choosing one chip over another. Samsung chips will vary this much among themselves as will the TSMC chips among themselves. Chips that fall outside this tolerance specification are not used in the phone. This is how all chip manufacturing works.

All this will be understood by those in the community with any manufacturing experience for any product or part. No two chips or phones can ever be indentical. The variation between chips and phones is within the manufacturing tolerances. And will always be seen in artificial tests that take a component to its maximum capacity.

Trading in a phone for a different chip is a fools choice. It will be different for the geekbench test only. And by the way the geekbench test can not even repeat. You won't get the same results running test again on same phone. It too has variability.

With all this said, some people will continue with irrational behavior based on either false information, repeated rumors taken as fact, or a lack of understanding for the significance or lack there of of tests. These individuals will pursue beliefs and feeling rather than facts. In many cases it will be ironic indeed when the phone they switch out for will actual be inferior in some other way. This is a lottery, a manufacturing lottery we all play with everything we buy.

 

[phpmaven]

Looking at the web all over the world, US, Japan, China, Taiwan, Europe which in discussion about this issue, this has been statistically significant that TSMC A9 is 20-30% more power efficient than the Samsung. I'd bet the sample size is way over 30. You can't call it not significant. So apple has to do something to compensate those who purchased iPhone 6s or 6s plus with Samsung chip. No matter what Apple say about the 2-3% differences between normal people's normal usage of their phone. It's not going to comfort people who care about it.

What do you think if you buy a new car with different outsourced parts ( I believe this is common in car industry), and you found your car engine is 20-30% less fuel efficient than another same new car when you NEED the power (he, when you accelerate or climbing). Will you claim to exchange one or report to government agency for investigation? Let's say your normal person?

If Apple makes mistake, we should let the company know that they should do something. A lot of people ignore that fact but claiming themselves not feeling any difference, please think about how much money you paid and how much money apple made on these phones especially they out-sourced to two companies risk themselves for the cpu only wants to cost down by purchasing chips from two competing companies.

Be good to human being.

Apple "has" to do something? I assure you they don't and they won't.

 

[MaciMac100]

At the end of video results are,

Samsung 55%
TSMC 62%

If we then keep on going, and do the math;

Samsung 10%
TSMC 24%

I would call this significant difference in battery life.

 

[terraphantm]

Exactly!
The test is flawed because the smaller Samsung chip running continuously heats up more and consequently uses (wastes) more power. The larger TSMC chip which runs a very slight bit slower and is larger surface area provides slightly more efficient cooling and does not get as warm. This means less power waste from tunneling.

The test is extremely artificial subjecting the chip to extreme loads it does not see in everyday use of any kind. Chips are normally seeing bursts of use with resting cycles. This means that the heat build up seen in the test and consequent increased energy consumption is not experienced outside of the test.

A truly relevant test would be to look at several thousands of phones under various conditions, locations, screen brightness, etc. tabulate a statistical mean and standard deviation and produce a six sigma bell curve of total battery durations. Do this for each chip phone calls mbination and compare those bell curves. Now you would have a meaningful test result to make the determination of which chip is superior in daily use.

Without this kind of data, the tests on your part with individual phones are completely irrelevant and lead to totally false assumptions as they have for the OP and some others.

The only party in a position to gather such data is Apple itself via the user experience data download it does from the each phone daily. Unless you have turned this feature off. This is how Apple arrives at the 2-3% difference it has stated publicly.

Since this is also within the statistical variation between the same chips made from silicone wafers. There is NO significance to choosing one chip over another. Samsung chips will vary this much among themselves as will the TSMC chips among themselves. Chips that fall outside this tolerance specification are not used in the phone. This is how all chip manufacturing works.

All this will be understood by those in the community with any manufacturing experience for any product or part. No two chips or phones can ever be indentical. The variation between chips and phones is within the manufacturing tolerances. And will always be seen in artificial tests that take a component to its maximum capacity.

Trading in a phone for a different chip is a fools choice. It will be different for the geekbench test only. And by the way the geekbench test can not even repeat. You won't get the same results running test again on same phone. It too has variability.

With all this said, some people will continue with irrational behavior based on either false information, repeated rumors taken as fact, or a lack of understanding for the significance or lack there of of tests. These individuals will pursue beliefs and feeling rather than facts. In many cases it will be ironic indeed when the phone they switch out for will actual be inferior in some other way. This is a lottery, a manufacturing lottery we all play with everything we buy.

Denial ain't just a river in Egypt.

 

[HEK]

At the end of video results are,

Samsung 55%
TSMC 62%

If we then keep on going, and do the math;

Samsung 10%
TSMC 24%

I would call this significant difference in battery life.

You are still making same mistake as others testing only two phones are making. There is no way you can be sure that the screen for instance on both phones is drawing the same current. Even if you set them to same percentage you still are assuming same current draw. The screens as with all other parts also have manufacturing variability. The battery itself as well are not identical. And the percent readout given by the phone has been notoriously inaccurate.

These tests need to be done on at least 300 phones each to even begin to draw any realistic conclusions.

 

[terraphantm]

You are still making same mistake others testing only two phones ar making. There is no way you can be sure that the screen for instance on both phones is drawing the same current. Even if you set them to same percentage you still are assuming same current draw. The screens as w lomas all other parts also have manufacturing variability. The battery itself as well are not identical.

These tests need to be done on at least 300 phones each to even begin to draw any realistic conclusions.

That argument loses weight as more and more people repeat the tests and have the same results. And you can have good data with a sample smaller than 300

 

[HEK]

Denial ain't just a river in Egypt.

Applying same statistical analysis techniques as i do at work. If I drew conclusions from data so far presented, I would be fired.

I am concluding nothing but that the data presented is inconclusive. And jumping to conclusion that you have inferior phone is way premature. But if you wanna jump go right ahead. Makes no never mind to me. LMAO

 

[raab]

Aren't you annoyed that you swapped out the Samsung chip which performs slightly better than the TSMC?

So now you have a slightly lesser performing phone for the sake of slightly better battery.

Just find it amusing, and sad, the lengths some people are going to.

 

[Mlrollin91]

That argument loses weight as more and more people repeat the tests and have the same results. And you can have good data with a sample smaller than 300

It doesn't matter if more and more tests have the same result. Until all variables are eliminated the final results aren't valid.

 

[th3r3ds0x]

Here is my take. For someone that uses the phone to play games which does in fact hit the cpu it's a pretty huge deal. For the person that doesn't do cpu intensive tasks on their phone then enjoy your Samsung processors. I like to know that I could potentially see an extra hour or more of game time from my phone with my TSMC A9. I think this may turn out to be a bit of a problem for Apple and we may see them quietly stop using the Samsung chips.

 

[Prince134]

Exactly!
The test is flawed because the smaller Samsung chip running continuously heats up more and consequently uses (wastes) more power. The larger TSMC chip which runs a very slight bit slower and is larger surface area provides slightly more efficient cooling and does not get as warm. This means less power waste from tunneling.

The test is extremely artificial subjecting the chip to extreme loads it does not see in everyday use of any kind. Chips are normally seeing bursts of use with resting cycles. This means that the heat build up seen in the test and consequent increased energy consumption is not experienced outside of the test.

A truly relevant test would be to look at several thousands of phones under various conditions, locations, screen brightness, etc. tabulate a statistical mean and standard deviation and produce a six sigma bell curve of total battery durations. Do this for each chip phone calls mbination and compare those bell curves. Now you would have a meaningful test result to make the determination of which chip is superior in daily use.

Without this kind of data, the tests on your part with individual phones are completely irrelevant and lead to totally false assumptions as they have for the OP and some others.

The only party in a position to gather such data is Apple itself via the user experience data download it does from the each phone daily. Unless you have turned this feature off. This is how Apple arrives at the 2-3% difference it has stated publicly.

Since this is also within the statistical variation between the same chips made from silicone wafers. There is NO significance to choosing one chip over another. Samsung chips will vary this much among themselves as will the TSMC chips among themselves. Chips that fall outside this tolerance specification are not used in the phone. This is how all chip manufacturing works.

All this will be understood by those in the community with any manufacturing experience for any product or part. No two chips or phones can ever be indentical. The variation between chips and phones is within the manufacturing tolerances. And will always be seen in artificial tests that take a component to its maximum capacity.

Trading in a phone for a different chip is a fools choice. It will be different for the geekbench test only. And by the way the geekbench test can not even repeat. You won't get the same results running test again on same phone. It too has variability.

With all this said, some people will continue with irrational behavior based on either false information, repeated rumors taken as fact, or a lack of understanding for the significance or lack there of of tests. These individuals will pursue beliefs and feeling rather than facts. In many cases it will be ironic indeed when the phone they switch out for will actual be inferior in some other way. This is a lottery, a manufacturing lottery we all play with everything we buy.

I looks like you are knowledgeable but you choose to deny it.
What you said variation in Samsung chips is correct. What you said variation among TSMC chips is correct.
What people said here is the two CPU comparison. It's already significant. You can't deny it.
(same as Volkswagon TDI car has emmision variation amongst them, but what if people test the car with Macedes diesel?). The truth is that the power efficiency between the A9 chips made by two company is obvious. Think about how Intel's CPU performance/ power efficiency with 30% difference can cost you? I don't know who you are. But you point out some people have irrational behavior when knowing the 20-30% power efficiency, making your article sounds trying to cover APPLE's responsibility.

 

[HEK]

That argument loses weight as more and more people repeat the tests and have the same results. And you can have good data with a sample smaller than 300

Repeating same flawed experiment doesn't prove anything except what I stated, that running chips balls out does cause extra battery drain on Samsung chip. If I that's sufficient info for you wonderful.

We have easily seen that much battery variability just from location of phone and radio reception, even from day to day. In real world use there are so many variables to battery consumption a sample size of 300 is barely acceptable. Typically one try's to reduce as many variables in an experiment to isolate causality.

This may be a hard concept for many of to understand. So go ahead and leap to conclusions. Run the same test over and over again. In fact, if that is what you use your phone for. Running the geekbench test. You definitely should exchange till you have the TSMC chip.

I just honestly don't see this as something to stress over in my so far quite enjoyable use of the phone.

 

[HEK]

I looks like you are knowledgeable but you choose to deny it.
What you said variation in Samsung chips is correct. What you said variation among TSMC chips is correct.
What people said here is the two CPU comparison. It's already significant. You can't deny it.
(same as Volkswagon TDI car has emmision variation amongst them, but what if people test the car with Macedes diesel?). The truth is that the power efficiency between the A9 chips made by two company is obvious. Think about how Intel's CPU performance/ power efficiency with 30% difference can cost you? I don't know who you are. But you point out some people have irrational behavior when knowing the 20-30% power efficiency, making your article sounds trying to cover APPLE's responsibility.

If I have two engines and run them balls out on test stand to failure and one lasts 20% longer than the other. Is one really better than the other when placed in a passenger car that neither engine sees that kind of loading and both engines work perfectly fine for life of car?

I'm not boosting or rationalizing nor making excuses for Apple. Many things they do like skirt taxes in US by incorporating in Ireleand I find reprehensible even if legal. I just don't subscribe to the whole sale stampede of feeling that you have an inferior product because of the chip you randomly received.

I believe most people will be happy with their phone till the 7 comes out regardless of which chip they have. Because in day to day real world use this whole thing amounts to a tempest in a tea cup. Only those continuously running this test running the chip to overheat conditions will be affected to any realistic meaningful user experience.

 

[5105973]

I do not have either 6s or 6plus. But I believe the thing we are discussing is a fact rather than some blog's bias opinion. No matter if you are APPLE fan or not you should react correctly, i.e. what's good to you, and APPLE. I think APPLE should compensate consumer but not trying to cover it. It sounds some people here are helping APPLE to cover it...

Compensate me for what? I get a full day of heavy use on a charge. The only time I don't is if I have to be on FB or FB Messenger with one of my friends or relatives for a few hours for some wackadoodle reason, like the other day some dude scared my friend by threatening on her FB page to off himself and I had to discuss it with her. Good God I hate FB.

Anyway, in any other use on any other App that is actually well written and not a steaming pile of resource hoggery, I get plenty of battery life. I actually had to go out of my way to make an effort to drain it the first day.

It's the best iPhone I've ever owned. It's got a beautiful white balance to the display. Call clarity is perfection. It's a pleasure to use. What exactly does IPhone owe me beyond that?

The only offer of improvement I'd take them up on is the camera software. I think they could fine tune their noise reduction algorithm a bit. That's all.

I've read and read and read about this subject until my eyes have crossed to make sure my best interests are not getting screwed over with this Chipgate stuff, meaning I'm not getting cheated out of promised performance. From everything I've read there currently is no scientifically, statistically sound evidence that having a Samsung A9 vs TSMC is a true detriment to my usage pattern. If that changes, if ironclad evidence demonstrates the TSMC really does give two hours extra battery life under the circumstances I'm likely to operate under, I'll be envious but go on with my life. I know my iPhone is meeting the promised specs. I'm not entitled to more than that.

 

[Aston441]

I'm fed up with Apple hiding specifications. Would you buy a computer with unknown specifications? The specs should be on the box and I should be able to buy what I want for my $1000.

MacBooks have different price points for different processors. Why should a handheld computer be different? If Apple wants to put different performance processors in their phones, fine, they can. But they need to be up front about that and price accordingly, with the lesser performance phone costing less. If it costs Apple money for me to repeatedly return my phone until I get the one I am trying to buy with my$1000, that is their fault, not mine.

 

[Aston441]

Also look for Apple to worsen battery life on TSMC phones in 9.03 to match Samsung phones to try to head off a law suit.

 

[roeiz]

Also look for Apple to worsen battery life on TSMC phones in 9.03 to match Samsung phones to try to head off a law suit.

hee hee hee, this is gonna be good...

 

[HEK]

Compensate me for what? I get a full day of heavy use on a charge. The only time I don't is if I have to be on FB or FB Messenger with one of my friends or relatives for a few hours for some wackadoodle reason, like the other day some dude scared my friend by threatening on her FB page to off himself and I had to discuss it with her. Good God I hate FB.

Anyway, in any other use on any other App that is actually well written and not a steaming pile of resource hoggery, I get plenty of battery life. I actually had to go out of my way to make an effort to drain it the first day.

It's the best iPhone I've ever owned. It's got a beautiful white balance to the display. Call clarity is perfection. It's a pleasure to use. What exactly does IPhone owe me beyond that?

The only offer of improvement I'd take them up on is the camera software. I think they could fine tune their noise reduction algorithm a bit. That's all.

I've read and read and read about this subject until my eyes have crossed to make sure my best interests are not getting screwed over with this Chipgate stuff, meaning I'm not getting cheated out of promised performance. From everything I've read there currently is no scientifically, statistically sound evidence that having a Samsung A9 vs TSMC is a true detriment to my usage pattern. If that changes, if ironclad evidence demonstrates the TSMC really does give two hours extra battery life under the circumstances I'm likely to operate under, I'll be envious but go on with my life. I know my iPhone is meeting the promised specs. I'm not entitled to more than that.

Very well said. You express my sentiments exactly. I'm pleased as punch with my 6s plus and have NO issues with battery life or the phenomenal speed of this phone. It more than meets my need 100% of the time. In truth I am amazed I am pulling what amounts to a MacBook out of my pocket.

 

[HEK]

I'm fed up with Apple hiding specifications. Would you buy a computer with unknown specifications? The specs should be on the box and I should be able to buy what I want for my $1000.

MacBooks have different price points for different processors. Why should a handheld computer be different? If Apple wants to put different performance processors in their phones, fine, they can. But they need to be up front about that and price accordingly, with the lesser performance phone costing less. If it costs Apple money for me to repeatedly return my phone until I get the one I am trying to buy with my$1000, that is their fault, not mine.

So which you want the faster Samsung processor for $50 extra or the slower more battery efficient TSMC processor for $50 extra. Your choice for only slight additional cost. DON't give Apple any ideas! Phone expensive enough already. Jeeeez

 

[HEK]

Also look for Apple to worsen battery life on TSMC phones in 9.03 to match Samsung phones to try to head off a law suit.

Go ahead and sue. Can see discussion with lawyer now ....... So where does Apple make a claim about the product they didn't meet?

Dah, well this guy on macrumers said my chip ain't as good. See he ran this geek test.

Get the hell out of my office! Case closed

 

[Gooberton]

This is a serious issue, wtf are you haters talking about. Samsung chip is much worse stop justifying and sticking up for Apple, return your phones till your happy

 

[Prince134]

I'm fed up with Apple hiding specifications. Would you buy a computer with unknown specifications? The specs should be on the box and I should be able to buy what I want for my $1000.

MacBooks have different price points for different processors. Why should a handheld computer be different? If Apple wants to put different performance processors in their phones, fine, they can. But they need to be up front about that and price accordingly, with the lesser performance phone costing less. If it costs Apple money for me to repeatedly return my phone until I get the one I am trying to buy with my$1000, that is their fault, not mine.

Those who say "normally you can't find difference" is insane if those did get an S chip iPhone. Think if they've wanted to get a macbook or macbook pro, that's several hundreds dollars difference because of different CPU to the basic. Will just try to ask APPLE to sell him Macbook Pro at same price as macbook, because as APPLE says normal people in a normal life experience they can't tell the difference. Does it sound right?

 

[bruinsrme]

Also look for Apple to worsen battery life on TSMC phones in 9.03 to match Samsung phones to try to head off a law suit.

Lawsuit for what?
On what grounds would a suit be justified?

Look at Apple site. The spec sheet clearly states the battery to last up to.
It doesn't say it will. What it clearly says is;
Internet use up to 12 hours.
Not 12 hours but rather but up to 12 hours.

Along with that is this caveat.

1. All battery claims depend on network configuration and many other factors; actual results will vary. Battery has limited recharge cycles and may eventually need to be replaced by Apple service provider. Battery life and charge cycles vary by use and settings. See www.apple.com/batteries and www.apple.com/iphone/battery.html for more information

Many other factors. Hmmm. I am sure different components will fall under that.

Apple will quietly tweak something to make the Samsung perform better.
I don't see a lawsuit here but YMMV

 

[Gooberton]

I'm fed up with Apple hiding specifications. Would you buy a computer with unknown specifications? The specs should be on the box and I should be able to buy what I want for my $1000.

MacBooks have different price points for different processors. Why should a handheld computer be different? If Apple wants to put different performance processors in their phones, fine, they can. But they need to be up front about that and price accordingly, with the lesser performance phone costing less. If it costs Apple money for me to repeatedly return my phone until I get the one I am trying to buy with my$1000, that is their fault, not mine.

This couldn't be more spot on! F u Apple. You want my 1,005.00 dollars, I'm gonna be picky. You can't even tell me the amount of ram because you don't want a spec war, well I wanna know

 

[HEK]

This couldn't be more spot on! F u Apple. You want my 1,005.00 dollars, I'm gonna be picky. You can't even tell me the amount of ram because you don't want a spec war, well I wanna know

Go buy a sammysung. Best specs ever written on paper.