GPU Performance

The 4th generation iPad integrates a quad-core PowerVR SGX 554 (MP4). The 554MP4 doubles USSE2 count over the previous generation PowerVR SGX 543MP4 used in the iPad 3, while keeping ROP and TMU counts the same. The result is a pure doubling of peak theoretical shader performance:

Mobile SoC GPU Comparison
  PowerVR SGX 543 PowerVR SGX 543MP2 PowerVR SGX 543MP3 PowerVR SGX 543MP4 PowerVR SGX 554 PowerVR SGX 554MP2 PowerVR SGX 554MP4
Used In - iPad 2/mini iPhone 5 iPad 3 - - iPad 4
SIMD Name USSE2 USSE2 USSE2 USSE2 USSE2 USSE2 USSE2
# of SIMDs 4 8 12 16 8 16 32
MADs per SIMD 4 4 4 4 4 4 4
Total MADs 16 32 48 64 32 64 128
GFLOPS @ 300MHz 9.6 GFLOPS 19.2 GFLOPS 28.8 GFLOPS 38.4 GFLOPS 19.2 GFLOPS 38.4 GFLOPS 76.8 GFLOPS

The theoretical numbers validate Apple's "2x faster GPU" claims, but as always we'll turn to Kishonti's GLBenchmark to see how achievable that performance increase is.

We'll start out with the raw theoretical numbers beginning with fill rate:

GLBenchmark 2.5 - Fill Test

GLBenchmark 2.5 - Fill Test (Offscreen 1080p)

The peak fill rate test shows a ~16% increase in performance over the previous generation 543MP4. Since there's no increase in number of TMUs we're seeing the results of a higher clocked GPU in the iPad 4's A6X.

GLBenchmark 2.5 - Triangle Texture Test

GLBenchmark 2.5 - Triangle Texture Test (Offscreen 1080p)

There's a pretty hefty improvement in triangle throughput - we're seeing more than a 60% gain compared to the iPad 3.

GLBenchmark 2.5 - Triangle Texture Test - Fragment Lit

At native resolution the fragment lit triangle texture test shows a big gain over the iPad 3 (~80%).

GLBenchmark 2.5 - Triangle Texture Test - Fragment Lit (Offscreen 1080p)

In both of the final triangle throughput tests the iPad 4 manages a 40 - 45% increase in performance over the iPad 3:

GLBenchmark 2.5 - Triangle Texture Test - Vertex Lit

GLBenchmark 2.5 - Triangle Texture Test - Vertex Lit (Offscreen 1080p)

With the synthetics out of the way, we can look at simulated game performance using the Egypt HD and Egypt Classic benchmarks. Remember the on-screen tests are run at native resolution with v-sync enabled, while the offscreen tests are run at 1080p with v-sync disabled for an architectural apples-to-apples comparison.

GLBenchmark 2.5 - Egypt HD

Despite sub-2x gains in a lot of the synthetic tests, Egypt HD shows us what's possible in a simulated game: the new iPad is roughly twice the speed of the previous gen model when running at the panel's native resolution. How we've seen this implemented in many cases is with titles finally running at native resolution on the iPad 4 vs. some lower, scaled resolution on the iPad 3.

GLBenchmark 2.5 - Egypt HD (Offscreen 1080p)

The Egypt Classic test is a much lighter workload, as a result most of these devices hit 60 fps at their native resolution:

GLBenchmark 2.5 - Egypt Classic

Although Egypt HD is a bit overkill for today's games, Classic undershoots by a good amount. The offscreen test however does provide some guidance as to whether or not these devices would be able to hit 30 fps on an appreciably heavier workload:

GLBenchmark 2.5 - Egypt Classic (Offscreen 1080p)

CPU Performance & Memory Bandwidth PowerVR SGX 554MP4 and iPad Retina Display: A Balanced Platform
POST A COMMENT

113 Comments

View All Comments

  • seapeople - Thursday, December 06, 2012 - link

    How late would this review have had to be for you not to post that comment? In fact, I bet I you typed that comment up the moment you heard about the Ipad release, saved it in a Word document, and patiently waited to post it on the Ipad review that would obviously be coming.

    The fact that Anand is practically the last tech site on the planet to actually post a review of the new Ipad makes it extremely ironic that you use this action to insinuate that Apple is paying Anand off.

    Oh well, it makes sense considering you can't even spell your own name right.
    Reply
  • bplewis24 - Thursday, December 06, 2012 - link

    It would be great to see the same image/resolution/detail comparisons that you did with the iPad iterations done with the iPad4 vs the Nexus 10 when it's ready. Reply
  • Zink - Thursday, December 06, 2012 - link

    I agree. You don't need to do it with every new tablet but it is a good opportunity to do some really in depth head to head iOS vs Android on 10" tablets. We all know this iPad is basically like the last one. What is harder to understand and get data on is completely different products. Reply
  • daar - Thursday, December 06, 2012 - link

    Would be nice to see in future reviews, can help get an idea of quality control and calibration, despite what Apple guarantees. Reply
  • Zink - Thursday, December 06, 2012 - link

    "I no longer have a 45nm iPad 2 so forgive me for the incomplete dataset here (it's safe to say that the iPad 2 would at least equal the iPad 4 in battery life, if not exceed it)"

    That doesn't seem safe at all from the numbers you gave. The 45nm iPad 2 only hits 10.1 hours in the 4th gen battery rundown and it is going to take a big drop with the new test. The iPad 3 sees a 12% hit to battery life with the switch to the new smartphone web browsing battery life test. Both tests should use the same screen settings so the increased energy consumption is coming from the 45nm silicon and radio. Screen res and GPU energy consumption does not significant contribute to this decrease in runtime because it is a browser test. The iPad 4 is also able to improve on the newer test despite its high res. Crunching the numbers, the iPad 3 is using 600 mW more to run the new test. The iPad 2 uses a bit older tech inside so I would assume it needs a similar increase in energy, putting the 45nm iPad 2’s performance in the new battery life test at around 8.2 hours, possible worse.
    Reply
  • Zink - Thursday, December 06, 2012 - link

    I didn't see you had the iPad mini numbers there too. Even it sees a small hit on the new test so the iPad 3 GPU isn't what is hitting battery life the hardest. I would be surprised if the 45nm iPad 2 can do even 8.5 hours, assuming these battery rundown tests are fairly accurate. Reply
  • Velius - Friday, December 07, 2012 - link

    Author states, "I've viewed the divide there as a line separating a computing device from an appliance. There's overlap in capabilities, but the overall experience tends to fall along those lines for me."

    This is a false meme that keeps getting propagated. Android is just as much of a walled garden as iOS, and I'm not talking about the App Store. I mean the functionality of the OS. Android is usually touted as having a lot more customizability than iOS, but this is an illusion. Sure, there are context-sensitive menus which are nice and all, but what you can and cannot do with them are a function of the app maker. It's like having a choice between Coke and Pepsi: you think you have a choice, but your options have been severely constrained. I've used both iOS and Android, the latter mainly to see what all the hype was about (as I am a Linux geek too, though prefer OS X for day-to-day), yet was astounded and shocked when I eventually tried it (on a SGS3 which I am now selling). Where was the full customizability? Where was this amazingly flexible system with tons and tons of features, bells, whistles, and doodads? It's just a copy of iOS! You have apps, and you run one at a time. That's just about it!

    As in many other things, Apple leads, and the rest (Google, Samsung, et al.) follow. Everyone forgets that when the iPad was first rumored, speculations were rife with regards to the capabilities of its OS. Many in the tech world wanted a full-blown OS X experience on the iPad (the name wasn't revealed back then of course). Yet they got the "appliance" that iOS on the iPad is. And yet, instead of forging its own path by introducing a full-blown Linux system on a tablet or phone, Google chose to copy and reproduce the appliance aspect of iOS in Android. Only now is Canonical actually moving to port Ubuntu to phones/tablets, well it's really an Android/Ubuntu hybrid but you actually get a full desktop when you dock it. *That* I'm looking forward to.

    So Android is just as much of an appliance as iOS, and iOS is just as much of a computing device as Android, if not more so. Yes, I'm talking about the superiority of the App Store - both in terms of quantity and quality. I have never seen a "killer app" in the Play Store that is also not present in the App Store (or with a better version). But ooh boy is it the case the other way around. There are some specialized scientific apps I use for research that just plain don't exist in Android. E.g., X-Windows Phase Plane (XPP) on the iPad:

    https://itunes.apple.com/us/app/xpp/id433859546?mt...

    Furthermore, of the apps that do get released for both Android and iOS, they generally work better in iOS. These considerations are in fact the main reason why I stick to Apple. The Nexus 10 et al. may be sexier in terms of hardware, but without good software, they are just not as functional.

    The level of Android fanboyism on the Internet truly astounds me. The gall, to claim that Android is a "computing device", and iOS an "appliance"! Like it's Cray supercomputer vs. a toaster! How astoundingly ridiculous! Unfortunately, even though this meme is groundless, it will keep getting propagated, ironically mostly due to the tech/geek "cred" that Android fanboys confer onto themselves, failing to realize that they are not the only ones with such cred. Heck, iOS fanboys are also Linux users. (You're looking at one right here.) I wonder, though, how many Android fanboys are *really* computer savvy?

    Comments like the ones the author made, and that inspired this post, make me wonder otherwise.
    Reply
  • Velius - Friday, December 07, 2012 - link

    Sorry about the vitriolic nature of that post. I'd erase and restart if I could.

    All I wanted to say is that modern tablets are pretty uniform in their capabilities and interfaces, so "computing device" vs "appliance" is a false distinction. Either they are all appliances, or all computing devices. (I see it as both.)

    In any case, great review. I will echo what someone else said, that it's a real treat to have a review of the iPad 4 in particular especially as it's not that different than the iPad 3.

    Keep up the good work. :)
    Reply
  • DeciusStrabo - Friday, December 07, 2012 - link

    What exactly are you missing in terms of customization of Android? I mean, you can go and switch out basically everything you want, Kernel, drivers, UI, functionality. You can add any software you want from any point you want (alternative App shops, your own developers, your own Linux/OS X machine). You can add your own codecs, drivers, functions. You can automate basically any part of the system if you so desire. So what are you missing exactly? Reply
  • Sufo - Friday, December 07, 2012 - link

    I have to agree with DeciusStrabo. You really aren't getting the most out of your android device if you can't see the distinction between iOS and Android (this coming from someone who thinks both are pretty shitty). There are many many things you can do with Android _devices_ that you can't do with an iphone. Try flashing custom roms for example. You _can_ change pretty much every aspect of an Android device - assuming you can root it. Can you run full OS VMs in iOS? I don't know, but I know you can with Android. In terms of apps, mxplayer is better than any iOS video player, and poweramp is better than any audio player - I'd say these are pretty good wins (at the same time, iOS default audio player shits on the Android default). That's not to say I disagree with you - iOS certainly has the edge often (app-wise), but your conclusions about Android are unfair and not accurate.

    So yes, a stock manufacturer Android experience is much closer to iOS than many people would like to admit (though it *is still* more of a computing device for sure). However if you know how to leverage the most out of an Android device, then there are certain powerful things it can do than an iOS device simply can't. It's essentially trading killer apps for specific device use cases, which leaves the two pretty equal overall, but with a clear favourite for most people depending on what you want out of your device.
    Reply

Log in

Don't have an account? Sign up now