Tests show differences between A9 chips from new iPhones; know exactly what we're talking about [atualizado 3x]

Tests show differences between A9 chips from new iPhones; know exactly what we're talking about [atualizado 3x]

Ah what is an iPhone release without a controversy, anyway? The #antennagate and #bendgate are to prove that many people love to search and find defects in Apple devices. It is important to note that, yes, the signal / reception of the iPhone 4 decreases if the user's hand is “covering” part of the antenna (in the device's frame). And yes, some iPhones 6/6 Plus units would warp more easily if a reasonable force was applied to a specific point on the device.

We can even call this a design flaw (after all, the iPhone 4s didn’t come with a “problem” in its antenna, just like the iPhones 6s / 6s Plus had their housing greatly reinforced making it impossible to bend under normal conditions), but the truth that many, but many other devices from famous manufacturers had the same problems and were not “charged” by the media in the same way. It turns out: the spotlight is always on Apple, whether for good or for evil.

Colorful 6s iPhones flying

The new iPhones could not do without their dose of polemic, and the subject of choice this time was the processor A9. In fact the processors as we reported, Apple is using two different models of chips, one manufactured by Samsung and another for TSMC. The big difference between them is that the one manufactured by the South Korean is a little smaller (about 10%), which generated some doubts: whether they both use the same manufacturing process (16 nanometers) or Samsung is using a more advanced technology advanced (14 nanometers)? Does this differentiation in size (and, who knows, in the manufacturing process) influence the performance of the chip? The most energy efficient Samsung processor, speaking?

Yes, they use different manufacturing processes. Yes, there are differences in processing performance and energy efficiency, but, unlike what was imagined, that of TSMC that presented the best results. More on that below.

Apparently, as we also talked about, there is no rule in the distribution of chips. Regardless of the device you have purchased (6s or 6s Plus), it can come equipped with any of them and there is no way to know this before purchasing the iPhone. Recent research by a developer has shown that in the 6s model the predominance of the A9 manufactured by TSMC, while in the 6s Plus of the Samsung processor.

Using two or more suppliers for the same component is quite normal in the world of technology, but when it shows different results (in this case, in performance and energy efficiency), it gets a little more complicated. And from what we are seeing, this is the case.

Tests done by YouTubers found that TSMC's A9, although larger, has a better energy efficiency than Samsung's processor. See s:

The first test done by Jonathan Morrison was to do a time-lapse 30 minutes on two devices with different chips, both with 100% battery and adjusted exactly the same. The result:

  • iPhone 6s with Samsung processor: 84% battery
  • iPhone 6s with TSMC processor: 89% battery

Then, Morrison made a 10 minute video in 4K resolution. After concluded, the batteries of both were like this:

  • iPhone 6s with Samsung processor: 75% battery
  • iPhone 6s with TSMC processor: 80% battery

Morrison then took that 10-minute video and exported it to iMovie to find out if there is any difference in performance between the processors. Samsung's A9 completed the task 4 seconds before TSMC's A9. The batteries looked like this:

  • iPhone 6s with Samsung processor: 60% battery
  • iPhone 6s with TSMC processor: 66% battery

When running the Geekbench 3 app on iPhones, Morrison noticed that the processors have quite similar performance in the test single-core, but that of TSMC did a little better in the multi-core (which is strange considering that Samsung's A9 was able to finish exporting the video 4 seconds earlier). Finally, Morrison made graphic tests (using the Metal environment) and found that both ran exactly the same, at 39.9 frames per second. The batteries after all these tests looked like this:

  • iPhone 6s with Samsung processor: 55% battery
  • iPhone 6s with TSMC processor: 62% battery

J Austin Evans performed deeper tests, such as the battery offered by the Geekbench 3 app.

Evans adjusted both iPhones to the same screen brightness and ran the test until both reached 50% battery life. Because TSMC's A9 lasted 50 minutes longer than Samsung's! That same test showed that while the TSMC chip temperature was 30C, Samsung's was 33C.

The user raydizzle, from reddit, also did that same battery test from Geekbench 3, but until the level reaches 0%. Here is the result:

Battery test - A9 processorsTSMC left; Samsung right

As we can see, TSMC's A9 lasted almost 2 hours longer than Samsung's processor!

In a more “real” test, Evans played a YouTube video on both devices (set in the same way), for 1 hour. While TSMC's consumed 14% of battery, Samsung's consumed 15%. Quite different from more laboratory tests, right? That is, in a scenario of normal use of the device the processors behave in a very similar way; but in a scenario of more extreme use, that of TSMC does a little better, yes.

We previously informed you that there was a way to find out if the chip on your iPhone 6s / 6s Plus from Samsung or TSMC. However, as it is an unreliable method, we do not recommend it. Now there is another, much simpler and more reliable way to kill your curiosity. Just install the Lirum Device Info Lite app and look for the chip reference. If it is N71AP (6s) or N66AP (6s Plus), from Samsung; if it is N71mAP (6s) or N66mAP (6s Plus), from TSMC.

Interestingly, the app went down a little while ago, let's hope it's something momentary and that everything will return to normal soon.

I honestly don't think users should be concerned if they discover they have the Samsung chip in their devices. In more real tests, as we have seen, the difference is not as stark as we are promoted by the Geekbench 3 app. But that Apple could have avoided this media headache more, it could

(via 9to5Mac, Engadget)

Update · 10/8/2015 s 13:20

Lirum Labs reported that it removed the app from the store due to compatibility issues with the new devices, and with iOS 9. It will be unavailable until a new update is sent to the App Store.

(tip from Andr Dal Molin)

Update II, by Rafael Fischmann · 10/8/2015 s 17:36

Apple has now given a statement on the matter to the TechCrunch, stating that real use tests with the devices show a variation of no more than 2-3% on battery consumption:

With the A9 chip designed by Apple on your iPhone 6s or iPhone 6s Plus, you are getting the most advanced smartphone chip in the world. Each chip we ship meets Apple's highest standards by providing incredible performance and delivering optimal battery life, regardless of capacity, color or model of iPhone 6s.

Some manufactured lab tests that make processors work with a continuous heavy load until the battery runs out are not representative of actual use, as they spend unrealistic time in the highest possible CPU performance state. a misleading way of measuring actual battery life. Our tests and consumer data show that the battery life of the iPhone 6s and iPhone 6s Plus, even taking into account differences in components, varies by only 2-3%.

In short, Apple is not saying that the tests carried out so far are liars, but that they are not a good indication of the daily use of the devices. If we take her data into account, this 2-3% variation is so small that it can occur even on technically identical devices.

Update III, by Rafael Fischmann · 10/8/2015 s 18:12

And the app we mentioned in the post came back from the App Store, without being updated.

Lirum Device Info app icon