Apple iPhone 6S battery life may be better if its uses TSMC A9 SoC

[HEXUS]

TSMC sourced A9 processor runs cooler, offers better battery life in stress test benchmark.

Read more.

[Kumagawa]

If this turns out to be representative rather than just an outlier due to binning (for either or both) then it's the opposite of what most people's responses on various forums have been.

[jinjur]

soooooo....its 50:50 as to whether you get better battery life?

[LSG501]

I'd maybe trust the post on reddit if the testing was 'exactly' the same and had performance figures too. Not to mention there are other hardware variables which might be at play such as the battery, other 'control chips' etc...
They're reporting the Samsung runs hotter so it could be thermal throttling (less cpu performance) playing a part too while Samsungs isn't and maintains full performance.

Also just purely from a 'business perspective' the odds of Apple allowing Samsungs cpu to perform 2 hours differently in exactly the same testing scenario is pretty slim, it would cause no end of issues for Apple in terms of advertising of stated performance/usage hours if nothing else. Not to mention the chip is 'designed' by Apple these days.

[SvDKILLSWITCH]

If this turns out to be representative rather than just an outlier due to binning (for either or both) then it's the opposite of what most people's responses on various forums have been.

I'm hardly an expert, but I feel this could be put down to the A9 chip having one overall design, and that design playing nicer with TSMC's 16FF process as opposed to Samsung's 14FF process.

It is kind of odd though. You'd expect a smaller process to use less power and be more efficient, but perhaps the design of the A9 is the limiting factor here in that it just suits the 16FF process better. Like I said though, I'm not an expert, just some guy commenting on an article.

[bob3406]

Or as a conspiracy theory, maybe Samsung has deliberately hamstrung the A9 on their 14nm process to screw with Apple...

[zsde]

Or as a conspiracy theory, maybe Samsung has deliberately hamstrung the A9 on their 14nm process to screw with Apple...

I doubt that. There would be a SLA in place and the parts supplied would have to comply with a minimum requirements specification. If they didn't, then Apple would reject them during testing. One can only assume they did comply with requirements but that the TSCM ended up exceeding those minimum requirements by a huge margin.

[zsde]

Or as a conspiracy theory, maybe Samsung has deliberately hamstrung the A9 on their 14nm process to screw with Apple...

Here is a counter conspiracy. Maybe Apple runs them at a higher clock or voltage to actually make them look bad, then claims to be a victim and in the process achieves the goal of making Samsung look bad? The cost of that may justify the ends? Not dissimilar to the VW test software manipulation? LOL

[peterb]

Ooh, a conspiracy...





Oh yes, forgot this is the interweb!

[DanceswithUnix]

I doubt that. There would be a SLA in place and the parts supplied would have to comply with a minimum requirements specification. If they didn't, then Apple would reject them during testing. One can only assume they did comply with requirements but that the TSCM ended up exceeding those minimum requirements by a huge margin.

Ah but maybe they only work OK during testing, and ... Nah, no-one would be that stupid

[watercooled]

Since when has a sample size of one been statistically significant and newsworthy? Until I see something to the contrary, I'm going with random variance (though such a large difference would seem unlikely) or more likely either a fault or nonidentical testing methodology.

[mtyson]

Update: Apple has uncharacteristically made a statement about the difference between the A9 processors:
"With the Apple-designed A9 chip in your iPhone 6s or iPhone 6s Plus, you are getting the most advanced smartphone chip in the world. Every chip we ship meets Apple's highest standards for providing incredible performance and deliver great battery life, regardless of iPhone 6s capacity, color, or model.

Certain manufactured lab tests which run the processors with a continuous heavy workload until the battery depletes are not representative of real-world usage, since they spend an unrealistic amount of time at the highest CPU performance state. It's a misleading way to measure real-world battery life. Our testing and customer data show the actual battery life of the iPhone 6s and iPhone 6s Plus, even taking into account variable component differences, vary within just 2-3% of each other."

Meanwhile more people are doing tests to see if there are differences in real-world battery depletion rates.

[LSG501]

Apple has uncharacteristically made a statement

um... they make a statement EVERY time there's an 'issue' which could affect their cash cow...

[watercooled]

It does seem fairly consistent in the side-by-side tests (though it's worth noting that isn't in any way proof - it's quite possible the TSMC>Samsung reports are just getting more attention for instance), however I wonder if it's actually down to the processor itself? It's not uncommon for Apple to multi-source multiple components in their devices e.g. NAND/RAM and many other parts.

It would be interesting to see a more thorough investigation to figure out where the differences are, preferably with a larger, random sample group.