You wouldn’t think it should matter, but several published reports claim that the battery life of an iPhone 6s and an iPhone 6s Plus can vary drastically depending on who built the processor. So Apple uses parts from TSMC (for Taiwan Semiconductor Manufacturing Company) and Samsung. The story goes that, if you get a handset with the former, you can get up to two hours additional battery life. These figures were based on a Geekbench 3 test that promises to measure expected battery life.
If true, and based on a possible breakdown of 50% for each manufacturer, you theoretically have a one in two chance of getting an iPhone with the right processor. But Apple has already stepped into the fray, asserting that the tests were wrong, that, in fact, the battery life difference is closer to two or three percent. That could still represent the difference between running out of power and having a few more minutes before the unit blacks out, but it’s otherwise not significant.
So just what is going on here?
Well, the folks at Ars Technica decided to run their own tests using four different measurement methods. The results were published on Monday morning, and, for the most part, it appears Apple is correct. Well, at least for the Wi-Fi browsing, WebGL and GFXBench tests. For the WebGL results, it appears the Samsung chip was marginally better.
But, as reported elsewhere, the Samsung chip falls way behind in the Geekbench 3 test.
Ars Technica is being realistic, and points out that such tests have limitations, and you can’t make a blanket assumption based on how the one iPhone fares against another in one set of canned tests. It may exacerbate the minor differences between the two chips, which actually benchmark about the same otherwise.
Obviously, there’s no guarantee which chip you’ll receive when you buy an iPhone. You can check the model numbers that appears when you run a benchmark, but otherwise, it’s not that you can demand that your Apple Store give you one with an A9 processor fabricated by TSMC. They’d laugh at you, or try to deal with such “unreasonable” requests with logic and reason.
So does Apple have a serious problem here? Will people with iPhones equipped with the “wrong” chip demand a replacement? It’s not something that a traditional manufacturer’s limited warranty would address, since Apple can no doubt assert that the product meets their specifications, and, besides, they reserve the right to buy parts from different sources.
Besides, has any other smartphone ever been checked for a similar phenomenon? Last year’s iPhone 6 and iPhone 6 Plus contained processors made by either TSMC or Samsung. Did anybody bother to measure the battery life of each to see if it was any different? What about other handset makers who source processors and other parts from different vendors? Or is Apple’s new hardware the only gear that has been subjected to this sort of benchmark?
Besides, when you use a canned benchmark, you are bound by its test algorithms, and if there’s something wrong in the programming that exaggerates differences under certain test conditions, that’s not something for Apple to fix. I’m not saying that the test is wrong, but it’s curious that only one type of test reveals this problem. I also expect that other publications will join Ars Technica in running their own battery life tests to see what sort of results they get. If the results confirm a problem without Geekbench 3, that may be one thing. If not, maybe we should blame the test.
Regardless, I’m sure some lawyers around the world are just salivating over the prospects of filing class action lawsuits against Apple asserting a product defect. But would different levels of power efficiency in parts that are otherwise compatible constitute a defect, or just a normal variation? What’s more, is it just the processor, or how it functions with other parts that Apple uses that may cause such a symptom to reveal itself?
I’m just guessing. I have no idea whether this problem exists or not, but it’s not unusual for a scandal to erupt involving Apple. Last year we had “BendGate,” the false claim that the iPhone 6 Plus was unduly sensitive to bending when placed in such locations as one’s rear pocket. It doesn’t matter that independent tests, including those run by Consumer Reports, failed to reveal evidence of any unusual sensitivity to damage under normal use and service. The story never died.
It was even resurrected this year in connection with the news that Apple used 7000 series aluminum for the new models. They are said to be far more resistant to bending, but was Apple responding to a reliability issue, or just doing their due diligence in making products better?
So will “chipgate” have any traction? I’m not at all certain, although the possibility will be sharply diminished if further tests, with different benchmarking tools, fail to confirm what the Geekbench 3 tests reveal. But that won’t stop some people from complaining.