The GPU

3D rendering is a massively parallel problem. Your GPU ultimately has to determine the color value of each pixel which may not remain constant between frames, at a rate of dozens of times per second. The iPad 2 had 786,432 pixels in its display, and by all available measures its GPU was more than sufficient to drive that resolution. The new iPad has 3.14 million pixels to drive. The iPad 2's GPU would not be sufficient.

When we first heard Apple using the term A5X to refer to the new iPad's SoC, I assumed we were looking at a die shrunk, higher clock version of the A5. As soon as it became evident that Apple remained on Samsung's 45nm LP process, higher clocks were out of the question. The only room for improving performance was to go wider. Thankfully, as 3D rendering is a massively parallel problem, simply adding more GPU execution resources tends to be a great way of dealing with a more complex workload. The iPad 2 shocked the world with its dual-core PowerVR SGX 543MP2 GPU, and the 3rd generation iPad doubled the amount of execution hardware with its quad-core PowerVR SGX 543MP4.

Mobile SoC GPU Comparison
  Adreno 225 PowerVR SGX 540 PowerVR SGX 543MP2 PowerVR SGX 543MP4 Mali-400 MP4 Tegra 2 Tegra 3
SIMD Name - USSE USSE2 USSE2 Core Core Core
# of SIMDs 8 4 8 16 4 + 1 8 12
MADs per SIMD 4 2 4 4 4 / 2 1 1
Total MADs 32 8 32 64 18 8 12
GFLOPS @ 200MHz 12.8 GFLOPS 3.2 GFLOPS 12.8 GFLOPS 25.6 GFLOPS 7.2 GFLOPS 3.2 GFLOPS 4.8 GFLOPS
GFLOPS @ 300MHz 19.2  GFLOPS 4.8 GFLOPS 19.2 GFLOPS 38.4
GFLOPS
10.8 GFLOPS 4.8 GFLOPS 7.2 GFLOPS
GFLOPS As Shipped by Apple/ASUS - - 16 GFLOPS 32 GFLOPS - - 12
GFLOPS

We see this approach all of the time in desktop and notebook GPUs. To allow games to run at higher resolutions, companies like AMD and NVIDIA simply build bigger GPUs. These bigger GPUs have more execution resources and typically more memory bandwidth, which allows them to handle rendering to higher resolution displays.

Apple acted no differently than a GPU company would in this case. When faced with the challenge of rendering to a 3.14MP display, Apple increased compute horsepower and memory bandwidth. What's surprising about Apple's move is that the A5X isn't a $600 desktop GPU, it's a sub 4W mobile SoC. And did I mention that Apple isn't a GPU company?

That's quite possibly the most impressive part of all of this. Apple isn't a GPU company. It's a customer of GPU companies like AMD and NVIDIA, yet Apple has done what even NVIDIA would not do: commit to building an SoC with an insanely powerful GPU.

I whipped up an image to help illustrate. Below is a representation, to-scale, of Apple and NVIDIA SoCs, their die size, and time of first product introduction:

If we look back to NVIDIA's Tegra 2, it wasn't a bad SoC—it was basically identical in size to Apple's A4. The problem was that the Tegra 2 made its debut a full year after Apple's A4 did. The more appropriate comparison would be between the Tegra 2 and the A5, both of which were in products in the first half of 2011. Apple's A5 was nearly 2.5x the size of NVIDIA's Tegra 2. A good hunk of that added die area came from the A5's GPU. Tegra 3 took a step in the right direction but once again, at 80mm^2 the A5 was still over 50% larger.

The A5X obviously dwarfs everything, at around twice the size of NVIDIA's Tegra 3 and 33.6% larger than Apple's A5. With silicon, size isn't everything, but when we're talking about similar architectures on similar manufacturing processes, size does matter. Apple has been consistently outspending NVIDIA when it comes to silicon area, resulting in a raw horsepower advantage, which in turns results in better peak GPU performance.

Apple Builds a Quad-Channel (128-bit) Memory Controller

There's another side effect that you get by having a huge die: room for wide memory interfaces. Silicon layout is a balancing act. You want density to lower costs, but you don't want hotspots so you need heavy compute logic to be spread out. You want wide IO interfaces but you don't want them to be too wide because then you'll cause your die area to balloon as a result. There's only so much room on the perimeter of your SoC to get data out of the chip, hence the close relationship between die size and interface width.

Most mobile SoCs are equipped with either a single or dual-channel LP-DDR2 memory controller. Unlike in the desktop/notebook space where a single DDR2/DDR3 channel refers to a 64-bit wide interface, in the mobile SoC world a single channel is 32-bits wide. Both Qualcomm and NVIDIA use single-channel interfaces, with Snapdragon S4 finally making the jump to dual-channel this year. Apple, Samsung, and TI have used dual-channel LP-DDR2 interfaces instead.

With the A5X Apple did the unthinkable and outfitted the chip with four 32-bit wide LP-DDR2 memory controllers. The confirmation comes from two separate sources. First we have the annotated A5X floorplan courtesy of UBMTechInsights:

You can see the four DDR interfaces around the lower edge of the SoC. Secondly, we have the part numbers of the discrete DRAM devices on the opposite side of the motherboard. Chipworks and iFixit played the DRAM lottery and won samples with both Samsung and Elpida LP-DDR2 devices on-board, respectively. While both Samsung and Elpida do a bad job of updating public part number decoders, both strings match up very closely to 216-ball PoP 2x32-bit PoP DRAM devices. The part numbers don't match up exactly, but they are close enough that I believe we're simply looking at a discrete flavor of those PoP DRAM devices.


K3PE4E400M-XG is the Samsung part number for a 2x32-bit LPDDR2 device, K3PE4E400E-XG is the part used in the iPad. I've made bold the only difference.

A cross reference with JEDEC's LP-DDR2 spec tells us that there is an official spec for a single package, 216-ball dual-channel (2x32-bit) LP-DDR2 device, likely what's used here on the new iPad.


The ball out for a 216-ball, single-package, dual-channel (64-bit) LPDDR2 DRAM

This gives the A5X a 128-bit wide memory interface, double what the closest competition can muster and putting it on par with what we've come to expect from modern x86 CPUs and mainstream GPUs.

The Geekbench memory tests show no improvement in bandwidth, which simply tells us that the interface from the CPU cores to the memory controller hasn't seen a similar increase in width.

Memory Bandwidth Comparison—Geekbench 2
  Apple iPad (3rd gen) ASUS TF Prime Apple iPad 2 Motorola Xyboard 10.1
Overall Memory Score 821 1079 829 1122
Read Sequential 312.0 MB/s 249.0 MB/s 347.1 MB/s 364.1 MB/s
Write Sequential 988.6 MB/s 1.33 GB/s 989.6 MB/s 1.32 GB/s
Stdlib Allocate 1.95 Mallocs/sec 2.25 Mallocs/sec 1.95 Mallocs/sec 2.2 Mallocs/sec
Stdlib Write 2.90 GB/s 1.82 GB/s 2.90 GB/s 1.97 GB/s
Stdlib Copy 554.6 MB/s 1.82 GB/s 564.5 MB/s 1.91 GB/s
Overall Stream Score 331 288 335 318
Stream Copy 456.4 MB/s 386.1 MB/s 466.6 MB/s 504 MB/s
Stream Scale 380.2 MB/s 351.9 MB/s 371.1 MB/s 478.5 MB/s
Stream Add 608.8 MB/s 446.8 MB/s 654.0 MB/s 420.1 MB/s
Stream Triad 457.7 MB/s 463.7 MB/s 437.1 MB/s 402.8 MB/s

Although Apple designed its own memory controller in the A5X, you can see that all of these A9 based SoCs deliver roughly similar memory performance. The numbers we're showing here aren't very good at all. Even though Geekbench has never been good at demonstrating peak memory controller efficiency to begin with, the Stream numbers are very bad. ARM's L2 cache controller is very limiting in the A9, something that should be addressed by the time the A15 rolls around.

Firing up the memory interface is a very costly action from a power standpoint, so it makes sense that Apple would only want to do so when absolutely necessary. Furthermore, notice how the memory interface moved from being closer to the CPU in A4/A5 to being adjacent to the GPU in the A5X. It would appear that only the GPU has access to all four channels.

The A5X SoC A Word on Packaging & Looking Forward
POST A COMMENT

232 Comments

View All Comments

  • zanon - Wednesday, March 28, 2012 - link

    In the article:
    Alternatively, we're used to a higher resolution enabling us to see more on a screen at one time. In the case of the new iPad, the higher resolution just makes things look sharper.

    The higher resolution does make smaller fonts readable. For something like an SSH session, that really will mean significantly more stuff can be on a screen at once.
    Reply
  • MobiusStrip - Thursday, March 29, 2012 - link

    A more useful change would be abandoning the ridiculous glossy screens. It's sad that Apple takes its cues from the plastic schlock being peddled at Best Buy, and participates in this fraud of shoving glossy screens down customers' throats. Reply
  • repoman27 - Thursday, March 29, 2012 - link

    The plastic schlock at Best Buy has a glossy plastic film applied to a cheap TN panel. Apple puts a piece of glass in front of their much more expensive IPS panels to protect them. The only way to make that glass (or the glass of the LCD panel itself) matte would be to apply an antiglare plastic film coating to the glass. These films have drawbacks (they block and scatter light making small details and text blurry.) The drawbacks become more exaggerated the farther the front surface of the glass is from the plane of the actual LCD.

    But you're right, it's probably Apple copying the design language of sub $500 laptops in order to somehow defraud the general public and force their customers to buy the products they actually produce.

    And seeing as how this discussion is about the new iPad screen, I'd like to point out that you're complaining about the lack of an antiglare coating on a touchscreen device... Strong work.
    Reply
  • Sabresiberian - Thursday, March 29, 2012 - link

    How is it fraud? Apple isn't, like, saying their screens are anti-reflective and then giving you totally reflective glossy screens.

    Many people prefer a glossy screen and simply aren't bothered by background reflections.

    ;)
    Reply
  • Henk Poley - Monday, April 02, 2012 - link

    Yes, Apple really should use Schott Conturan/Amiran/Mirogard antireflective technology.

    btw, not-glossy does not mean matte. Air is not matte either. Glass can be see-through too ;)
    Reply
  • Watwatwat - Monday, April 02, 2012 - link

    Nope, steve gibson has tested even using screen protectors on the new ipad vs not, it seems to affect the resolution at that level, matte might not be a good idea at all for high density display. Reply
  • KoolAidMan1 - Thursday, March 29, 2012 - link

    I wasn't initially blown away, but then after a day of using it every other display seemed bad in comparison. It is one of those things you didn't realize was needed until using it, now I want very high DPI in all of my monitors. Reply
  • menting - Wednesday, March 28, 2012 - link

    Is it just me, or do Shadowgun and GTA screenshots look more detailed in Transformer Prime than in the iPad? Reply
  • menting - Wednesday, March 28, 2012 - link

    nm..i just noticed that it's scaled up in new ipad, so it's definitely not as sharp.
    However, how can fps be fairly compared in this case then?
    Reply
  • TheJian - Wednesday, March 28, 2012 - link

    Basically because of the way Nvidia and Apple approach games so far, you can expect games on Tegra3 to just look better as they seem to aim for more graphics and fewer games (they spend money on fewer projects that produce better results), as opposed to apple who spreads the wealth but just ends up with more cannon fodder if you ask me :) You should get more variety on Apple I'd guess, but a better experience with fewer choices on Tegra3/Android. I like QUALITY over QUANTITY personally and hope Apple leans the way of Nvidia in the future. I would rather have 10 games that I'd play for weeks or months (if I'm playing on my hdtv through one of these I want better water, buildings etc) rather than games I fire up for less than 20 minutes as their just another angry birds variant and arguably useless on your TV.

    I want these devices to KILL the consoles next year and make MS/Nintendo etc give it up in 2015 or whenever the next revs should come. I hope they just realize we won't buy them anymore. DirectX11 on my phone/tablet and probably standard 25x14 resolutions by then (won't all be retina by 2015?) make a console purchase STUPID. This could be the merging of console/pc we need since phones/tablets rev yearly like pc's instead of 10yr console's stuck in stone stagnating gaming. Your phone as a portable console with xbox/ps3/pc gamepad support would be excellent. Pump it out to a monitor and keyboard/mouse setup and you have a notebook replacement too...LOL Now if they'd just put in a few extra cores by then that will disable if on their own screen but turn on when on a larger display like TV/Monitor and we have exactly what we want in both cases :)

    Pipe dreams? Retina is here now, and gamepads sort of. Next stop cores that only turn on depending on display output :) Awesome battery on the road, and great power in the dock at home pushing your 27in monitor. :) The 28nm versions by xmas of everyone's chips should come close to console power or surpass them. Interesting times.
    Reply

Log in

Don't have an account? Sign up now