TSMC chips can get 2.5hrs extra compared to Samsung on iPhone 6S Plus

[NEDM64]

Tests using battery % are not accurate.

Batteries vary from unit to unit, there are huge differences between them, and the % is an estimation, that can explain the difference.

Anandtech should make a real scientific test, with Wattmeters. Please!

 

[Midwayman]

Tests using battery % are not accurate.

Batteries vary from unit to unit, there are huge differences between them, and the % is an estimation, that can explain the difference.

Anandtech should make a real scientific test, with Wattmeters. Please!

Maybe, but its very consistent. I don't see anyone running the tests and saying samsung beat TSMC.

 

[Eug]

Apple says real world differences in battery life are 2-3%.

http://techcrunch.com/2015/10/08/ap...rmance-of-new-iphones-a9-chips-vary-only-2-3/

 

[mmntech]

Apple says real world differences in battery life are 2-3%.

http://techcrunch.com/2015/10/08/ap...rmance-of-new-iphones-a9-chips-vary-only-2-3/

Which is being backed up by users. The real world difference seems to fall around 6-11%.
http://www.macrumors.com/2015/10/08/samsung-tsmc-a9-videos/

I think for most people, that's not going to make much of a difference. There are are a select few users that will have a genuine gripe over losing a half hour battery life, but that's a tiny percentage of the user base.

I'm more curious to know what's causing this performance gap. The chips should be identical regardless of what fab they come out of. The issue only seems to affect the CPU and not the GPU.

 

[sm625]

I wonder if samsung gimped their chip intentionally just because apple is a competitor. Also, could it simply be the slightly higher temperature of the SoC that causes the extra power drain?

 

[Red Storm]

I wonder if samsung gimped their chip intentionally just because apple is a competitor. Also, could it simply be the slightly higher temperature of the SoC that causes the extra power drain?

That would be stupid by Samsung, and even so, it's not like they could sneak it in somehow. Apple verifies the product before it releases, so whatever discrepancy there is, Apple was definitely fully aware of it.

 

[davygee]

That would be stupid by Samsung, and even so, it's not like they could sneak it in somehow. Apple verifies the product before it releases, so whatever discrepancy there is, Apple was definitely fully aware of it.

As I posted in another thread, if Apple HAD to use both TSMC and Samsung to produce the chips for the phones, then why not have one manufacturers chip in the 6S and the other in the 6S Plus, this would have eliminated this whole issue.

 

[Eug]

As I posted in another thread, if Apple HAD to use both TSMC and Samsung to produce the chips for the phones, then why not have one manufacturers chip in the 6S and the other in the 6S Plus, this would have eliminated this whole issue.

Not really a viable option. The 6s sells in vastly greater volumes than the 6s Plus.

 

[Midwayman]

Apple says real world differences in battery life are 2-3%.

http://techcrunch.com/2015/10/08/ap...rmance-of-new-iphones-a9-chips-vary-only-2-3/

I would read that as "best case" is 2-3%.

 

[Eug]

I would read that as "best case" is 2-3%.

I wouldn't. I would read that as "2-3% difference for the average user".

 

[Midwayman]

I wouldn't. I would read that as "2-3% difference for the average user".

Apple's definition of average user is probably one quite favorable to them. They have a very vested interest in making that number look as low as possible.

 

[Oyeve]

Dorks, non-issue. Really, grow up.

 

[Printer Bandit]

are people losing sleep over this? like staying up at night thinking about it?

 

[Oyeve]

are people losing sleep over this? like staying up at night thinking about it?

Probably. I mean, if they have enough time to create elaborate blogs and testing images bitching about it.

 

[Midwayman]

You know if you're not interested in it you can just not read the thread rather than thread-crap.

 

[redgtxdi]

Interesting. I was talking yesterday with a customer about this. He had a 6 and now a 6s with the Sammy chip & is not liking the battery life he's getting. He'd typically finish a day between 20-25% battery & now finishing between 10-15% on average. (Yes, I'm aware 6s battery's smaller than 6, just saying)

I currently have a 6s + on backorder w/ Verizon & have noticed that suddenly everybody (including Apple) is calling for long deliveries on the 6s +. (Regular 6s appear to still be readily available)

I wonder if "chipgate" has anything to do with these long lead on the 6s+????

 

[gpse]

Interesting. I was talking yesterday with a customer about this. He had a 6 and now a 6s with the Sammy chip & is not liking the battery life he's getting. He'd typically finish a day between 20-25% battery & now finishing between 10-15% on average. (Yes, I'm aware 6s battery's smaller than 6, just saying)

I currently have a 6s + on backorder w/ Verizon & have noticed that suddenly everybody (including Apple) is calling for long deliveries on the 6s +. (Regular 6s appear to still be readily available)

I wonder if "chipgate" has anything to do with these long lead on the 6s+????

6s+ has been on backorder for a while now, They sold out very fast on the Apple Store when the pre-order begun. It has nothing to do with the Chips, they were having display issue's.
http://www.macrumors.com/2015/09/14/iphone-6s-short-supply-backlight-module-kuo/

 

[mmntech]

I wonder if samsung gimped their chip intentionally just because apple is a competitor. Also, could it simply be the slightly higher temperature of the SoC that causes the extra power drain?

Highly unlikely. Samsung's chip fab division is one of their top units for revenue right now, and a lot of that business is coming from the A9. If they intentionally gimped those chips, and Apple found out, they'd lose millions. They'd probably also get sued. You don't cut off your nose to spite your face.

 

[mmntech]

ArsTechnica ran their usual battery life test on both the Samsung and TSMC iPhones. Their findings match Apple's claim of a 2-3% difference for real world usage. The TSMC chip bested the Samsung by 14 minutes for WiFi web browsing, and 6min for GFXBench. Peanuts in the grand scheme of things.
http://arstechnica.com/apple/2015/10/samsung-vs-tsmc-comparing-the-battery-life-of-two-apple-a9s/

The Geekbench test again proved to be the only one with a significant difference. 28% in favour of the TSMC, or a full extra hour battery life. So it would see that the Samsung really doesn't like it when you max out CPU utilization. Either that or Geekbench's testing methodology is flawed.

 

[Midwayman]

The Geekbench test again proved to be the only one with a significant difference. 28% in favour of the TSMC, or a full extra hour battery life. So it would see that the Samsung really doesn't like it when you max out CPU utilization. Either that or Geekbench's testing methodology is flawed.

Geekbench is only like 30% utilization according to Anandtech. It would be interesting to track down exactly what is causing the issue though. Something CPU related. I'm wondering if Samsung has a higher TDP, but doesn't get to sleep as quickly. Something that is repeatedly pinging the CPU could be worst case with a slower time to sleep.

 

[redgtxdi]

Well, my delivery's been pushed back from 10/12 to 10/26. (sigh)

I have a feeling there's a little more going on than display supply. You can't even *select* a 6s Plus on Verizon (arguably the nation's largest cell provider) at Costco's website.

Interesting.

 

[schneiderguy]

As I posted in another thread, if Apple HAD to use both TSMC and Samsung to produce the chips for the phones, then why not have one manufacturers chip in the 6S and the other in the 6S Plus, this would have eliminated this whole issue.

At least for the vendors of the small ICs in their phones, Apple requires the ability to manufacture at two different fabs, presumably to reduce the risk of a supply interruption. So having a single source for the most important IC in their phone might be deemed as too risky.