Apple's A8-powered iPhone 6 & 6 Plus impress, outperform competing phones in benchmarks
The new silicon at the heart of Apple's recently-released iPhone 6 and iPhone 6 Plus has increased the company's lead in smartphone application processor performance, according to data released on Monday, with the A8 improving on the already industry-leading 64-bit A7.
The iPhone 6 saw an impressive 13 percent bump in performance over the iPhone 5s on the SunSpider browser performance benchmark, according to AnandTech. Other CPU-bound benchmarks followed a similar trajectory, with the iPhone 6 or 6 Plus leading the way in all but one category.
In light of the results, the publication reiterated its previous stance that the A8's performance improvements are unlikely to be the result of its 100 megahertz clock speed bump alone. At the time of the A8's announcement, the group wrote that Apple is likely to have "enhanced or otherwise significantly optimized" the A8's processor core, which would represent "a significant accomplishment" given the time frame involved.
Apple's latest handsets posted a more up-and-down showing when it came to GPU performance, finishing near the bottom of the pack in a physics simulation, while besting all but Nvidia's Shield tablet in most other areas. Overall, they showed "a pretty solid lead over the competition for the iPhone 6/A8."
One area of concern was the GPU performance of the iPhone 6 Plus. The 5.5-inch device --?which renders content at 2208-pixels-by-1242-pixels before downsampling to 1920-pixels-by-1080-pixels --?was more sluggish, which AnandTech posits is due to the required high-resolution rendering.
Finally, battery life was found to be "quite incredible." The iPhone 6's 11 hours of web browsing battery life was "a step above just about every other Android smartphone on the market," while the iPhone 6 Plus bested all but Huawei's 6.1-inch Ascend Mate 2 by at least two hours.
The iPhone 6 saw an impressive 13 percent bump in performance over the iPhone 5s on the SunSpider browser performance benchmark, according to AnandTech. Other CPU-bound benchmarks followed a similar trajectory, with the iPhone 6 or 6 Plus leading the way in all but one category.
In light of the results, the publication reiterated its previous stance that the A8's performance improvements are unlikely to be the result of its 100 megahertz clock speed bump alone. At the time of the A8's announcement, the group wrote that Apple is likely to have "enhanced or otherwise significantly optimized" the A8's processor core, which would represent "a significant accomplishment" given the time frame involved.
Apple's latest handsets posted a more up-and-down showing when it came to GPU performance, finishing near the bottom of the pack in a physics simulation, while besting all but Nvidia's Shield tablet in most other areas. Overall, they showed "a pretty solid lead over the competition for the iPhone 6/A8."
One area of concern was the GPU performance of the iPhone 6 Plus. The 5.5-inch device --?which renders content at 2208-pixels-by-1242-pixels before downsampling to 1920-pixels-by-1080-pixels --?was more sluggish, which AnandTech posits is due to the required high-resolution rendering.
Finally, battery life was found to be "quite incredible." The iPhone 6's 11 hours of web browsing battery life was "a step above just about every other Android smartphone on the market," while the iPhone 6 Plus bested all but Huawei's 6.1-inch Ascend Mate 2 by at least two hours.
Comments
Wait, why?
As for cellular speed, I see ping times that are about 2.5x slower with the 6 Plus than the year-old iPad Air (AT&T LTE, 3 bars). This is using Speedtest app. 160+ ms for the 6 Plus, ~60 ms for the iPad Air.
Wait, why?
Downsamples to render objects at the proper size for the screen. I would imagine a regular 1080p screen would greatly reduce GPU strain, but Apple may have considered improved sharpness and other LCD characteristics worth the tradeoff. See recent DisplayMate review.
Whats up here... With Metal , Scene and Sprite kit one would expect stellar performance !
My suspicion is that the benchmark is somehow not compatible with apples new technology ..
Any insights from those more in the know !?
Downsamples to render objects at the proper size for the screen. I would imagine a regular 1080p screen…
It HAS a regular 1080 screen. Why is it rendering above that resolution? Why is it downsampling at ALL?
As for cellular speed, I see ping times that are about 2.5x slower with the 6 Plus than the year-old iPad Air (AT&T LTE, 3 bars). This is using Speedtest app. 160+ ms for the 6 Plus, ~60 ms for the iPad Air.
and the actual upload and downnload speeds
Can anyone explain to me why web/javascript benchmarks are "a relatively useful proxy for CPU performance"? I just ran Octane, SunSpider, Kraken on my rMacBook (mid-2012) :
These are clearly not designed to be browser neutral and hardly represent a "relatively useful proxy" for anything, but web/Javascript performance. I guess it makes a lot of sense to compare the results within one product line, but these are far from being legit CPU benchmarks.
You're probably referring to the 3DMark result. That benchmark tests the CPU as well as the GPU--probably graphics-related tasks like physics simulations on the CPU. This may be one of those rare areas where having four cores actually improves benchmark scores, giving devices with more cores a slight advantage.
The 3d benchmark does not compute!
Whats up here... With Metal , Scene and Sprite kit one would expect stellar performance !
My suspicion is that the benchmark is somehow not compatible with apples new technology ..
Any insights from those more in the know !?
I'm not in the know... but some simple sleuthing points out that 3dMark V1.2 was last compiled for iOS 7.0, and the current version 1.3 was compiled for 7.1. My guess is there is no 'metal/scene/sprite' optimizations in the executable (calling the old frameworks).
The 3d benchmark does not compute!
Whats up here... With Metal , Scene and Sprite kit one would expect stellar performance !
My suspicion is that the benchmark is somehow not compatible with apples new technology ..
Any insights from those more in the know !?
I had the same inquiry. 3D Mark tests are based on "OpenGL ES 2.0." and the A8 is optimized for Metal, according to the keynote.
Is that how long it’s been since you’ve had your prescription changed?
The Phonearena battery test obviously uses a different methodology. While the 6 still does well, it is not ahead of the pack. I am not sure why anyone would expect the 6 and 6 plus to have extraordinary battery life. They are extraordinarily thin with a very powerful processor, it is unreasonable to expect those two features would deliver class leading battery life.
Wait, why?
http://www.paintcodeapp.com/news/iphone-6-screens-demystified
Wait! What???
Apple stuff "outperforms", in spite of not clearly winning those all-important spec wars,
and clearly copying Samsung's Noteworthy trailblazing???
I "carn't not believe it"!
( \SS.../:s...:;\...dang, I can never remember how to make that sarcasm thing...)
So why not just use fewer points in the initial design?
It would throw off display on the older, smaller phones. Everything is based on the resolution of the iPhone 5 screen (previous screens had the same horizontal resolution but different aspect ratio). Downsampling to a 1080p screen will result in much sharper images than downsampling to a lower resolution screen. Of course, upsampling is never good.