Implications of SGX543 in iPhone/Pod/Pad?

[Deleted member 13524]

SONY states MP4+ for one and on the other there's still the low level API advantage of a "closed environment" of a hand-held console amongst others.

I meant performance potential in the hardware, of course. That doesn't include possible low-level optimisations. And as far as we know, that "+" could mean something negligible.

But let's not forget that iOS can actually be considered a closed environment on its own. It's not like they'll have to develop drivers for a multitude of hardware combinations like Android (and WP7 in the mid-term future?).


That said, I doubt any developer will spend big loads of cash to develop AAA titles for the iPad2 anytime soon, like they will for the NGP, and that's where the low-level optimizations are made.

 

[Ailuros]

I meant performance potential in the hardware, of course. That doesn't include possible low-level optimisations. And as far as we know, that "+" could mean something negligible.

Depends on the perspective really, how "negligible" it actually is. Just for the record you can have all the hw potential in the world and the result could be underwhelming due to inferior sw/drivers.

That shouldn't mean that Apple's sw platform is inferior or any of the sorts but there are still quite a few differences between a hand-held console and a tablet or smart-phone.

But let's not forget that iOS can actually be considered a closed system on its own.

No it isn't compared in the sense it is on any (hand-held) console.

That said, I doubt any developer will spend big loads of cash to develop AAA titles for the iPad2 anytime soon, like they will for the NGP, and that's where the low-level optimizations are made.

An ISV could even develop an el-cheapo 5 bucks short title for the NGP and still have the low level API advantages. Oversimplified you are able to address the underlying hw more directly and I doubt Carmack is wrong when he claimed that it'll give NGP an advantage over an additional hw generation from the smart-phone realm.

 

[wco81]

Remember it will be out next week, while NGP won't be out for almost another year?

Shortly after the NGP releases, we could have SOCs which approach the NGP's profile.

 

[tangey]

Remember it will be out next week, while NGP won't be out for almost another year?

I think NGP is supposed to be OCT (7 months) ?

 

[Mike11]

I think NGP is supposed to be OCT (7 months) ?

Sometime in Q4, but AFAIK it's not clear yet if it's just gonna be Japan, US or worldwide.

 

[Lazy8s]

Though iOS standards are not as open as a PC's, they're still abstracted enough to allow direct and easy compatibility across multiple generations in contrast to dedicated platforms for dedicated consoles.

 

[ltcommander.data]

http://www.eetimes.com/electronics-news/4213873/Apple--TSMC-to-expand-foundry-ties

It's being reported now that Apple is going to move SoC production to TSMC including the A5 for the iPad 2. This is strange after the report that Apple quadrupled wafer allocation at Samsung. It'd make sense to do the next A6 on 28nm, but does Apple really need more A5 then the increased production at Samsung would provide? Moving the A5 to TSMC would no doubt require re-laying out the chip. Is a few nm or other differences in TSMC's 40nm process really that much of an improvement over Samsung's 45nm process? Unless one of these reports is wrong.

Although it could make sense that Apple produces the A5 for the iPad 2 and iPhone 5 at Samsung, along with the A4 for the iPhone 4 which looks to see continued strong demand with reports that the CDMA iPhone 4 is looking to launch in Asian countries later this year even after the iPhone 5 is released. The 5G iPod Touch, being the usual process guinea pig, could get the TSMC A5 along with a possible fall iPad 2.5 or iPad 3.

 

[ltcommander.data]

http://daringfireball.net/2011/03/the_ipad_2

The results show that the iPad 2’s graphics improvements far outshine its straightforward CPU improvements — exactly as Apple has advertised.

For example, on my original iPad, with 200 on-screen sprites, the framerate dropped to 45 fps. On the iPad 2, with 400 on-screen sprites, the framerate remained at 65 fps. On the iPad 1, Guy’s demo app dropped below 60 fps with about 100 animated sprites; on the iPad 2, it didn’t drop below 60 fps until there were over 750 animated sprites.

After I showed him the results, Guy told me, “The results show that the iPad 2 is easily about twice as powerful as the original and that this speed gain is a freebie — you don’t need to change your code structure in order to see significant gains. The differences in the amount of time spent rendering indicates that the GPU is really much faster than the original. The original iPad had a comparatively weak fill-rate and it was an issue for the device. The second generation really leaves that behind and it looks like it’ll be able to do some really incredible things graphically. My demo code is workman-like, competent code — meant to measure the relative strengths of the parts of the system. Taking some time to get the most out of that GPU and CPU will pay off with some really remarkable games and graphics apps.”

Gruber has some interesting testing on the relative graphics capabilities of the iPad and iPad 2. Looks like Apple's 9x raw graphics improvement gains are reasonable and the weak fill-rate issue has been resolved.

(Hopefully the long quote doesn't infringe on any copyrights.)

 

[Pressure]

Can't wait to see games that pushes the iPad 2.

It seems more and more like a powerhouse, especially when compared to other tablets.

 

[Lazy8s]

So, 512 MB confirmed then. Disappointing but still an impressive product overall.

 

[Pressure]

So, 512 MB confirmed then. Disappointing but still an impressive product overall.

Yeah, they need something for next years model too

 

[Ailuros]

Is there some sort of error in the GLBenchmark2.0 databank or what?

http://www.glbenchmark.com/phonedet...2&testgroup=overall&benchmark=glpro20&var=top

Egypt standard = 5113 frames (45.2fps)

If you now go to the compare page you'll get following results:

http://www.glbenchmark.com/compare....howhide=true&certified_only=1&D1=Apple iPad 2

Egypt standard = 6613 frames

Which would equal to 58-59fps where I'd frankly expect an iPad2 to be also considering it's vcynced at 60Hz.

Weird....

 

[Ailuros]

So that was it....http://www.glbenchmark.com/phonedetails.jsp?D=Apple+iPad+2&benchmark=glpro21

Anyone care to explain if there's any difference between 2.0 and 2.1?

Almost 10k frames with vsync off is damn impressive. Still not good enough

 

[Lazy8s]

Pixel warm up test is notably near the 250 MHz x 4 TMU theoretical. Like the Kyro benchmark days.

Nice to see it stretch its legs a little more in some of those tests.

 

[Ailuros]

What's a pixel warm up test anyway? (honest question)

 

[Lazy8s]

Sounds like it might be a little more synthetic than the other pixel and triangle fill rate tests it replaced in the other versions of GLBenchmark, or maybe just a name change. Not really sure how the test runs...

Maybe Laszlo will stop by.

 

[tangey]

A size-by-side compare with the Mali in the S2:-

http://www.glbenchmark.com/compare....D1=Apple iPad 2&D2=Samsung GT-i9100 Galaxy S2

SGX543x2 is killing the mali-400MP4 in virtually every individual test, many tests are x2, some are x3 or x4.

there are however two tests where the Mali is getting signficantly better results, both are Trig tests, with mali hit x2 and x3 the performance of the SGX543.

 

[Ailuros]

A size-by-side compare with the Mali in the S2:-

http://www.glbenchmark.com/compare.jsp?benchmark=glpro21&showhide=true&certified_only=2&D1=Apple%20iPad%202&D2=Samsung%20GT-i9100%20Galaxy%20S2

SGX543x2 is killing the mali-400MP4 in virtually every individual test, many tests are x2, some are x3 or x4.

there are however two tests where the Mali is getting signficantly better results, both are Trig tests, with mali hit x2 and x3 the performance of the SGX543.

The offscreen Egypt result for Exynos is quite low compared to the standard one and it's a phenomenon that appears on various so far tested platforms (like iPad1 and others). It's too early for comparisons IMO; I'm sure most if not all of them will be chasing after performance improvements for 2.1.

***edit: self correction: Kishonti just added a "720p" next to the offscreen in their results. Else its running when in off screen at 1280*720? Well that explains a lot then. Else for the moment irrelevant wether Hardkernel Odroid A or Galaxy S2 at offscreen 720p the iPad2 is a tad more than 2x times faster in Egypt or PRO.

 

[JohnH]

Finally, the offscreen tests should now allow some level of direct comparison between devices with different display resolutions and where vsync limits are kicking in....

 

[Lazy8s]

And a broad base of results are coming in quickly. Even a GC800 Vivante powered device, the Window N50, is represented, as well as a Mali 200 device, the Cat Stargate.

Though a myTouch 4G Slide, an Adreno 220 equipped phone, has been tested, it doesn't have an offscreen test result yet.

http://www.glbenchmark.com/result.j...3&os=4&version=all&certified_only=1&brand=all