The New iPhone 6s 'Chipgate' Stirs Battery Fears

There's a saying that what goes around comes around. Apple turned to two different companies to make its new A9 processor, and now the decision is backfiring.
As teardown experts Chipworks originally reported, Apple has two models of the CPU in the new iPhone 6s and 6s Plus: a smaller chip made by Samsung, and a slightly larger chip made by TSMC. The Samsung device is made on a 14nm process, while the TSMC chip is made on a 16nm process.
"For Apple to go through all the trouble of dual-sourcing a custom designed part and launching on day one with both parts, suggests major sourcing problems. For cost and power reasons, there is little reason to run a larger die, unless the smaller die was not available at the right volumes," Chipworks said.
This would all be just a curiosity except that Redditors started claiming, based on Geekbench battery tests done in China, that the TSMC chip has 20 percent better battery life than the Samsung one. The split between Samsung and TSMC chips isn't clear, with one person on the thread saying 25 percent of A9s are made by Samsung while another one says 60 percent are from Samsung.

The app Lirum Device Info ($2.99) supposedly explains what model of chip you have, although neither that app nor the competing iStat worked on our test iPhone 6s Plus$799.99 at Best Buy. So we don't know which kind of chip we have here at PCMag right now.
AnandTech has the best analysis of what's really going on here, but Apple weighed in with an official statement on the topic.
"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," Apple said. "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 percent of each other."

Battery Tests Are Hard
There are several levels to this debate.
The anxiety iPhone users are feeling starts with the fact that the iPhone 6s and 6s Plus simply don't have very good battery life. Apple has always privileged thinness over long usage time, and as Joanna Stern reported in the Wall Street Journal, battery life was the No. 1 user complaint about the previous iPhone. The new iPhones' batteries don't last as long as previous models, though to be fair, the Galaxy S6 doesn't have great battery life, either.
Battery testing is also extremely difficult, because there are a thousand ways people use their devices. We test devices using continuous LTE video streaming, for instance. That really pumps the radio and the screen but doesn't strain the processor. So a more power-efficient modem and screen would make more of a difference in our test than a power-efficient CPU. Geekbench pumps the CPU in ways that aren't totally realistic. Apple runs its tests with the screen brightness turned down. We turn it up. So the Redditors and Apple may both be right, just in different scenarios.
AnandTech pulls out another, lesser-known fact: there's also a wide variation in performance between chips coming off the same wafer, which is termed the "silicon lottery." So, just to make you more hysterical, the variation in performance between two Samsung chips or two TSMC chips may be greater than the variation between a Samsung chip and a TSMC chip. We just don't know.



"1-on-1 comparisons under controlled conditions can provide us with some insight in to how the TSMC and Samsung A9s compare, but due to the natural variation in chip quality, it's possible to end up testing two atypical phones and never know it," AnandTech says, correctly.
That means you should just stick with the iPhone you have, and understand there's always been a factor of luck in your device performance. You just never knew about it before.

Credit: Pcmag