Early Benchmarks confirm iPhone 5s’ GPU is twice as fast as iPhone 5


When Phil Schiller, Apple’s Senior Vice President of Worldwide Marketing unveiled the iPhone 5s at the media event on Tuesday, he told us that the next generation 64-bit A7 chip that powers it, comes with twice as fast CPU and twice as fast Graphical Processing Unit (GPU) compared to iPhone 5’s A6 chip.

The first benchmark results of iPhone 5s has just hit the web, which confirm that A7 chip indeed comes with twice as fast GPU.

Phone Arena reports that GFXBench 2.5 score of the iPhone 5s confirms that the graphics performance has improved considerably. They also point out since the GFXBench test does not support OPenGL|ES 3.0, which is supported by iPhone 5s’ A7 chip, we may not be seeing the full capabilities of Apple’s new system-on-chip.

In the 1080p off-screen test, the iPhone 5S shows nearly double the performance of its predecessor. This doesn’t show up in the on-screen test though as the benchmark hits the 60fps screen limit (the test is old and can’t stress the GPU at the native 640 x 1136 resolution of the iPhone 5s).

With OpenGL ES 3.0 support on both iOS 7 and Android 4.3 and improving mobile GPUs, we’re about to get a lot of eye candy on mobile 3D games.


We’ll have to wait and see how much more improvement support for OpenGL|ES 3.0 will have on the benchmark scores when the test is updated.

The faster CPU and GPU, along with support for OpenGL|ES 3.0, the latest graphics standard should significant improve the performance of graphics intense apps like games on the iPhone 5s.

Check out this post to find out everything we know about iPhone 5s’ A7 chip.

[Via PhoneArena]

Like this post? Share it!

  • richrich

    left s4 behind!

    • Alterian

      I’m good with my GS4. I had iPhone since they came out. Not any more there to simple and arent very customizable.
      A stock iPhone can’t compete with a GS4 but a JBroken iPhone is just as well if not better then a Galaxy. I can’t wait months for a jail break so I left that world behind. To each their own.

    • F*** dan

      Actually the lg g2 powers 4x more pixels than the iPhone 5. If the lg g2 had a 1136×640 display, then it and every 1080p device out there would destroy the iPhone 5s. Hate to ruin your false pride.

      • Richard Xing

        That doesn’t make a difference on the off-screen test where it still came second.

        • F*** dan

          The only way that would help is if you did everything off screen

        • clikes2004

          What is it doing off screen because if it is really running a higher resolution, it would naturally have a lower fps than at a lower resolution. This benchmark claims that it got faster off-screen which doesn’t make sense if it is doing 1080p. That would be like you only being able to lift 30 lbs 10 times and then being able to do 35 lbs 12 times. It doesn’t work that way. In fact, it’s impossible.

          • verdantchile

            The on-screen average is dragged down by the times when it’s capped by v-sync. So, despite off-screen having a higher resolution, the lack of v-sync when rendering off-screen produces a result that represents more of its potential performance.

          • Richard Xing

            It is rendering everything off-screen. The whole point of the off-screen test is to determine how far the GPU can go without wasting extra resources on the display.

  • Byron Swift

    I’m happy to see the improvements!!! It’s better than the Galaxy S4 but I think the LG2 and Note 3 might be faster since they have the S800. Keep in mind the screen on the S4, LG2, and Note 3 has much bigger screens.

    Does the screen size effect benchmarks?

    • JD

      its more the resolution than the size.

    • Richard Xing

      It doesn’t affect the off screen benchmark.

  • John

    this benchmarks are way off

  • Stephane M

    These numbers are designed to put the 5S forward… Just something to point out real quick, the first benchmark shows iPhone 5S on top, the iPad 4 is nowhere to be seen. And on the second one, it is there, want to make this a fair benchmark? Please add the resolutions next to the actual results.

    Also, when it comes to offscreen. Sure, it’s easy to say that the tool benchmarks in 1080p. But nowhere does it talks about smart API optimization to push smart upscaling with little to no taxation on the GPU (think Xbox 360 vs PS3). Mind you, a phone like for example the S4 or a tablet like iPad 4 runs by default on similar if not higher resolutions, therefore upscaling would not happen on the original benchmark. It would be pure results or more-so in the case of the iPad underscaling but then again, it’s nowhere near it’s default resolution OR default format which should cause to “tax” the GPU but would still logically net slightly higher results then say on-screen results.

    Last but not least, if you think I’m just saying whatever, please keep in mind that iPhone 5S’s resolution is 1136×640. Off-screen is 285% more pixels then native resolution… No real world test would give you 3 fps over if not for upscaling technology.