Microsoft Surface tablets

That didn't stop Apple from selling a gazillion iPad/iPad2 and soon to be decent sales for the iPad Mini. Let's not forget the 1280x800 screens on their MacBook Pro 13".

(BTW, I agree about the Kuro. Sad to see them go, but it seems being the best TV ever years before anyone caught up is not good enough to be profitable, as you need to leverage a halo position to move 10x as many lower end units. At least Panny has been able to bring that high contrast goodness down to the mainstream. I think I saw a 60" current gen panel going for $800 recently, and it just blows my mind that such a seemingly expensive technology can scale down to such low cost per square inch.)

 

I wonder what metric they used. For example, did they actually look for iPad user agents, or did they use "Mobile Safari" user agents?

Many Android browsers are based on WebKit, and as a result identify as Mobile Safari:
Mozilla/5.0 (Linux; U; Android 2.2; en-sa; HTC_DesireHD_A9191 Build/FRF91) AppleWebKit/533.1 (KHTML, like Gecko) Version/4.0 Mobile Safari/533.1

Of course, the user agent does contain device-specific information (see the bit in parenthesis, which says stuff about iPad on the iPad).

I'd like to know how and when Apple came up with that number.

 

That's a huge survey sample. Plenty to be accurate within a few % provided they use sound methodology.

 

I looked at anandtech but couldnt see any mention of it running portrait orientation, why omit mentioning it. Surely its possible? all other tablets have the option

 

What exactly would those x86 capabilities be?

Running legacy x86 software?

Most of the mobile software, designed for multitouch, is compiled for ARM.

 

Clovertrail is a disaster next to A6x.

Intel needs to land a slamdunk with it's Silvermont chip. While I expect Intel to bring it's A game with Airmont, it might be too late to win the volumes needed for it's precious fabs.

 

Don't know about that. They have an architecture license. They don't seem to be using it for consoles. Servers seem less likely than mobile chips. My guess is that they are building custom mobile chips to play Apple's game Apple's way some years down the line.

 

Thanks! The lack of user profiles in iPad has caused some frustration so this great. I don't mind having to purchase the apps multiple times if they're good.

 

The point was that there aren't any significant performance costs for AF for desktop games; mobile games are times more humble in about everything especially in terms of texture resolutions. If texture resolutions would in the majority of cases be as high as for the desktop we'd run into other headaches before AF like storage and or download bandwidth headaches.

Indirect under the reasoning that by the time you need to loop for any of the AF related calculations (due to the absence of quad TMUs for the lowest common denominators) the bandwidth requirements increase by far more.

Not a single disagreement there; before Apple even released the iPad3 when it became clear that it's going to have a 2048*1536 display I said that it would be far more reasonable to render any 3D in a much lower resolution, with the addition of MSAA/AF and can't think of a single reason why it would had been a problem. But that's still one exception since their tablet GPUs since the iPad2 have at least 4 TMUs. For all other cases like the ULP GF, Mali400MP1, SGX544MP1, Adreno2xx and the likes AF won't be an easy pill to swallow, despite that the majority of devices they're integrated in range from 720p to 1080p type of resolutions and not necessarily on as big screens.

The trend for high resolutions from smartphones to tablets isn't just an Apple exclusivity unfortunately. If I rate 2048 as overkill for a ~10" display, what should I rate 1080p on a 5" display instead? The positive thing about it is that I don't expect resolutions to scale beyond the current points because devices can't get any bigger than the already are. The next best target would be any of sort of connectivity to ultraHD devices but that's a different chapter.

 

We are very different. The 2048x1538 resolution of the iPad 3 is 80% of the reason I bought one. I waited particularly for one that had a resolution like this because I thought the 1024x768 was too grainy, especially for text consumption, which is a big part of the appeal.

Conversely with limited internal memory and frequent hardware updates leading to rapid fragmentation (lots of people bought an iPad over here, but I expect very few to want to upgrade more than every five years), the gaming aspect of this platform has, imho, a limited 'serious' future.

 

A6x will be slightly faster than Medfield in CPU performance, but I highly doubt it will be anywhere close to Clovertrail.

You can see the comparison between A6 (iPhone 5) and single core Medfield ATOM (RAZR i) here (http://www.anandtech.com/show/6386/samsung-galaxy-note-2-review-t-mobile-/3).

Medfield wins two CPU benchmarks, ties one, and loses one. Medfield has a single ATOM core, and it still can compete very well (win more benchmarks than lose) against the dual core A6. Clovertrail is a dual core version of Medfield. It will be significantly faster than Medfield. A6x on the other hand is only a slightly overclocked A6 (1.3 GHz -> 1.5 GHz if we believe the rumors). I don't see any possibility that A6x ends up being faster in CPU tasks than Clovertrail.

GPU performance is of course another matter. Clovertrail has a considerably faster GPU compared to Medfield (SGX 545 @ 533 MHz). It doubles the theoretical flops and peak geometry compared to Medfield. However A6 is already over twice as fast as Medfield, so I wouldn't expect Clovertrail to win this battle. But Clovertrail's GPU performance should be very much acceptable for 1366x768 screens.

For 1920x1080 screens we have Ivy Bridge with HD 4000 and it sports pretty much equal GPU performance compared to the current generation gaming consoles. Of course the downside being that you can run demanding games only for two hours or so, and the battery is dead

 

This advantage hardly matters in a tablet, because almost nobody runs computing-intensive apps on such a device. It kills the battery for starters, and second, even intel's portable chips are no damn good at computing anyway; not compared to a desktop processor.

Problem with HD4000 is it's an immediate-mode renderer, which spends a LOT of its fillrate and power budget drawing and shading invisible pixels. This is no good in a portable device - arguably no good in a desktop setting either really, but neither AMD nor Nvidia (nor Intel for that matter) seem interested in wanting to do anything real about this issue...yet, anyway.

Time and power constraints may change that eventually.

 

Ha we missed these biased Javascript/browser comparisons so much

What about some benchmark that doesn't rely on the effort put in the browser and the JS engine? For instance geekbench:
A6 vs Medfield 2GHz
Cortex-A9 vs Medfield 2GHz

Basically A6 already is faster than Medfield for single thread tasks and Clovertrail won't change that. Atom in its current form isn't competitive, its new core will surely change that, but we are at least one year from its release.

 

We might be not that different after all. Yes the iPad1 & 2 1024*768 resolution was way too low for the screen estate, but I could also argue that 2048*1536 is way too much for a sub-10" screen. What if they would had gone for 1600*1200 instead as just one example and simply throw in some font AA for text? Would you suggest that a good quality 13" laptop display with a way lower native resolution than even 1600 is at a serious disadvantage even if you zoom into text?

What happened to reasonable middle of the way solutions? It's not that going from one extreme to the other extreme is necessarily a wise decision. Even worse the upcoming Nexus 10 tablet will have a 2560*1600 display.

By that time we will have quad core Kraits and Swifts ready to launch in devices.

 

Not sure if that's what you meant, but you only get the former because of the latter. The degree of anisotropy at a sample point is not dependent on resolution (assuming constant aspect ratio).

iOS does use font AA, though it doesn't use subpixel rendering or the aggressive hinting ClearType uses. The problem with text rendering at lower resolutions is that you either get blurry text or you distort the glyph shapes and indeed the positioning within a line of text which means you can no longer zoom without causing formatting issues. It can be done pretty well, but nothing beats higher resolution.

For a device used primarily for reading text (which probably applies to many iPad users), the tradeoffs you have to make to get 250+ ppi are quite sensible.

 

You know they did it for app compatibility. But, no, it's not overkill IMO. It's simply beautiful, yet I still can see individual pixel on it when reading text, more so than for an iPhone. (I use it very often laying down, iPad resting in my rib cage, with screen much closer than when sitting upright.) With 16x12 it'd be much worse. Why not go for the best you can do, knowing very well that nobody won't be able to match for a long time, when the opportunity is there?

With iOS apps generally written to be fixed resolution, everything would look like crap for existing apps. Slightly higher weight is a decent trade-off.

Can't wait for the 3096x2048 version, though.

 

Xmas above just said that iOS actually uses font AA. Else the way I understand things its actually a combination of 2048*1536 with N samples of font AA. If I didn't understand him wrong that's something that can be beaten easily, since I was thinking all this time that iPad3 doesn't use any font AA at all.

 

If you are going to trash talk in browser benchmarks don't follow that up by using some closed source set of microbenchmarks with completely unknown compilers ...

PS. why is there no Coremark on the appstore?

 

I agree. My point is that it at least removes some dependencies on things you can't control. It's far from ideal, but it certainly is infinitely more pertinent than running some JS/browser benchmark to have some feeling about pure CPU performance.

Hmm there perhaps is some restriction on its use? Anyway I don't think Coremark is an interesting benchmark but at least the source is available

 

Android platform and software is mostly build on top of Java (Dalvik VM). Java benchmarks are very much relevant for Android. Only a small fraction of software is build on top of NDK.

If you want to benchmark native NDK software, you need to be aware, that the native software needs to be compiled separately for x86 target in order for it to run optimally on ATOM. If you run a native application that only has ARM version, Intel will perform ARM->x86 binary translation for the native machine code. This naturally results in lower performance. Are you sure the Geekbench isn't running on top of binary translation on ATOM, or have they made a separate x86 build for ATOM based Android devices?

iOS does font antialiasing. Otherwise all text would look very bad on iPad 2 (and other low DPI iOS devices). iOS however doesn't to do subpixel font antialiasing (it's not visible on screenshots / iPad 2 screen). Subpixel font antialiasing basically increases the screen (horizontal) resolution by 3x for (monochromatic) text rendering. It causes some color bleeding and requires exact knowledge of display subpixel physical layout. Screen rotation of course also requires the algorithm to change how it operates. So it requires OS and hardware level support to work properly.