iPhone 6s: Samsung and TSMC A9 chips yield different battery life
The Apple iPhone 6s/6s Plus
pair has been available for two weeks now and initial hype around its
new camera and 3D Touch is starting to wear off. Curious minds are now
digging deeper and are looking into the difference between the Samsung
and TSMC A9 chips in terms of performance and battery life.
The Samsung-made piece is manufactured on a 14nm production line, while TSMC employed a 16nm process. Teardowns revealed the former to measure 96 sq.mm, and the latter is a bit larger at 104.5 sq.mm. All that would lead you to believe the Samsung A9 to be more efficient, but in practice that appears not to be the case.
A couple of folks have rounded up a number of iPhones with different chips and done some comparison tests with results that are rather startling at first, but not all that disturbing in the end.
For example, running GeekBench parallel on two devices with identical settings shows the Samsung chip to be 33% less efficient than its TSMC counterpart. Or if you spin it the other way around, the TSMC will last up to 50% longer than the Samsung. In consecutive runs of GeekBench, that is, which is an isolated CPU benchmark and in no way representative of actual use. An hour-long YouTube video playback test results in 15% vs. 14% drop in battery level for the two respective chips, well within the margin of error for such a short test.
Another YouTuber has attempted a more life-like usage scenario for his comparison. The test kicks off with a 30-minute timelapse recording, which leaves the Samsung-equipped iPhone 6s at 84%, while the TSMC one is at 89%. Ten minutes of 4K video capture later, both phones are down an additional 9%, and an iMovie export of the same 4K video leaves the two devices at 60% vs. 66%.
It couldn’t go without some GeekBenching but in this case it was from a performance point of view and the TSMC model scores marginally higher. No difference is found between the two when subjected to a GPU-focused GFXBench Metal, and after the entire ordeal the iPhone 6s with a Samsung chip inside is left with 55% of juice, while the other one shows 62%.
There’s an app in the App Store called Lirum that can tell you the model of the chipset you got, but this can only happen post-purchase. In the end, there’s no way of knowing, which one you’re getting before you actually buy it. That said, unless your smartphone use is dominated by timelapse recordings (whose isn't), it shouldn’t make much difference for purposes practical.
Source 1 Source 2
The Samsung-made piece is manufactured on a 14nm production line, while TSMC employed a 16nm process. Teardowns revealed the former to measure 96 sq.mm, and the latter is a bit larger at 104.5 sq.mm. All that would lead you to believe the Samsung A9 to be more efficient, but in practice that appears not to be the case.
A couple of folks have rounded up a number of iPhones with different chips and done some comparison tests with results that are rather startling at first, but not all that disturbing in the end.
For example, running GeekBench parallel on two devices with identical settings shows the Samsung chip to be 33% less efficient than its TSMC counterpart. Or if you spin it the other way around, the TSMC will last up to 50% longer than the Samsung. In consecutive runs of GeekBench, that is, which is an isolated CPU benchmark and in no way representative of actual use. An hour-long YouTube video playback test results in 15% vs. 14% drop in battery level for the two respective chips, well within the margin of error for such a short test.
Another YouTuber has attempted a more life-like usage scenario for his comparison. The test kicks off with a 30-minute timelapse recording, which leaves the Samsung-equipped iPhone 6s at 84%, while the TSMC one is at 89%. Ten minutes of 4K video capture later, both phones are down an additional 9%, and an iMovie export of the same 4K video leaves the two devices at 60% vs. 66%.
It couldn’t go without some GeekBenching but in this case it was from a performance point of view and the TSMC model scores marginally higher. No difference is found between the two when subjected to a GPU-focused GFXBench Metal, and after the entire ordeal the iPhone 6s with a Samsung chip inside is left with 55% of juice, while the other one shows 62%.
There’s an app in the App Store called Lirum that can tell you the model of the chipset you got, but this can only happen post-purchase. In the end, there’s no way of knowing, which one you’re getting before you actually buy it. That said, unless your smartphone use is dominated by timelapse recordings (whose isn't), it shouldn’t make much difference for purposes practical.
Source 1 Source 2
Comments
Post a Comment
Kindly Comment Only related to Post