Sure I know there are other graphics benchmarks other than polys per sec however the article above only mentions polys and I was just pointing out jragosta's incorrect statement that the iPad 2's GPU wasn't any better than the 545.
In any case, real world performance of real apps on the actual devices (and their respective displays) is what it comes down to, whether it be GPU, CPU, memory speeds, etc.
The graphics benchmarks might be interesting from an egghead perspective, but they don't mean much to Joe Smartphone Consumer who doesn't look at specs.
That's kind of interesting you mention that. How great do the graphics have to be on a 4" iPhone and 4" iPod Touch. iPad 2 quality? Yeah, but that's about it for 2012, I reckon. The real battle for better graphics and performance will be the iPad 3. Because now you're taking on the "big boys" - laptops, desktops, gaming consoles, set-top boxes, everything. It's not just a post-PC device, it's a post-everything-digital kind of device.
That's the genius of Steve Jobs, always one step ahead. While smartphones still have a ways to evolve you can imagine some sort of saturation point until the next re-invention of the phone. But the iPad, that's fresh meat right there to dig into for a good few years... The scope is massive.
I think the value of Apple's acquisitions of P.A. Semi and Intrinsity will start to really show next year with the A6 SoC particularly in the iPad 3, especially if it moves to a Retina Display.
Could this have anything to do with the rumor that Apple was experimenting with some kind of dual e-ink/standard display to counter the "reading in bright sunlight" problem on the iPad?
Could this have anything to do with the rumor that Apple was experimenting with some kind of dual e-ink/standard display to counter the "reading in bright sunlight" problem on the iPad?
While the video claims to show a prototype iPhone, the hardware displayed looks identical to the currently available iPhone 4. It's possible that the graphics switching functionality, shown accessed through the iOS Settings application via the Developer Settings section, was enabled through a hack, or has been faked by some other method, such as a jailbreak.[...]]
...or even easier: develop a 'dummy' Settings App.
Honestly I can't see Apple sending out a machine to customers that had such a feature. So if real, this is likely a video of a device developers are working on at Apple.
As to the processor, there have been persistent rumors that Apple split the hardware development (SoC) teams in two. One team to develop chips for tablets and one team to develop for handheld devices like iPhone. This actually makes lots of sense from the hardware standpoint as space is far more limited in a phone as is power and performance expectations. On the flip side tablets need far more power to enable more advanced apps, drive high density displays and otherwise deliver a better user experience. More importantly tablets have more power available to them.
In the end the flick is interesting because it does indicate a processor split for iOS devices. Well it does if you expect iPad 3 to get a much more powerful GPU.
Could this have anything to do with the rumor that Apple was experimenting with some kind of dual e-ink/standard display to counter the "reading in bright sunlight" problem on the iPad?
Why? It appears to be faster than what the iPad 2 offers:
"But the A5 processor found in the iPad 2 has a SGX543 GPU, different from the "545" model shown in the video.
The SGX545 GPU was first unveiled in January of 2010, boasting full support for OpenGL 3.2 and OpenCL 1.0 with real-world performance of 40 million polygons per second at 200MHz. For comparison, the SGX535 can produce 28 million polygons per second, while the SGX543 pushes 35 million polygons per second."
But the 545 is the single core variant, the iPad 2 has the dual core variant of the 543, the SGX543MP2. Even if each unit is slower, the iPad 2 has...Well, two. The single core may be faster one on one, but the iPad 2's would still end up faster by a mile. And its not like SLI or Crossfire, the performance scaling is seamless. See here, the 545 isn't on there but you can extrapolate.
As for 32 cores, please note that in desktop computing, the trend of offering processors with more and more cores has decelerated. The same may well happen with mobile devices. Also, please note that we don't need 32 cores to do a billion polygons.
Just checking : )
i agree, which is why i said -if- cores continue. and no cpu cores do not maketh the poly count
Comments
Sure I know there are other graphics benchmarks other than polys per sec however the article above only mentions polys and I was just pointing out jragosta's incorrect statement that the iPad 2's GPU wasn't any better than the 545.
In any case, real world performance of real apps on the actual devices (and their respective displays) is what it comes down to, whether it be GPU, CPU, memory speeds, etc.
The graphics benchmarks might be interesting from an egghead perspective, but they don't mean much to Joe Smartphone Consumer who doesn't look at specs.
That's kind of interesting you mention that. How great do the graphics have to be on a 4" iPhone and 4" iPod Touch. iPad 2 quality? Yeah, but that's about it for 2012, I reckon. The real battle for better graphics and performance will be the iPad 3. Because now you're taking on the "big boys" - laptops, desktops, gaming consoles, set-top boxes, everything. It's not just a post-PC device, it's a post-everything-digital kind of device.
That's the genius of Steve Jobs, always one step ahead. While smartphones still have a ways to evolve you can imagine some sort of saturation point until the next re-invention of the phone. But the iPad, that's fresh meat right there to dig into for a good few years... The scope is massive.
Could this have anything to do with the rumor that Apple was experimenting with some kind of dual e-ink/standard display to counter the "reading in bright sunlight" problem on the iPad?
No.
[...]
While the video claims to show a prototype iPhone, the hardware displayed looks identical to the currently available iPhone 4. It's possible that the graphics switching functionality, shown accessed through the iOS Settings application via the Developer Settings section, was enabled through a hack, or has been faked by some other method, such as a jailbreak.[...]]
...or even easier: develop a 'dummy' Settings App.
No.
Thanks for clearing that up.
As to the processor, there have been persistent rumors that Apple split the hardware development (SoC) teams in two. One team to develop chips for tablets and one team to develop for handheld devices like iPhone. This actually makes lots of sense from the hardware standpoint as space is far more limited in a phone as is power and performance expectations. On the flip side tablets need far more power to enable more advanced apps, drive high density displays and otherwise deliver a better user experience. More importantly tablets have more power available to them.
In the end the flick is interesting because it does indicate a processor split for iOS devices. Well it does if you expect iPad 3 to get a much more powerful GPU.
Could this have anything to do with the rumor that Apple was experimenting with some kind of dual e-ink/standard display to counter the "reading in bright sunlight" problem on the iPad?
No.
Why? It appears to be faster than what the iPad 2 offers:
"But the A5 processor found in the iPad 2 has a SGX543 GPU, different from the "545" model shown in the video.
The SGX545 GPU was first unveiled in January of 2010, boasting full support for OpenGL 3.2 and OpenCL 1.0 with real-world performance of 40 million polygons per second at 200MHz. For comparison, the SGX535 can produce 28 million polygons per second, while the SGX543 pushes 35 million polygons per second."
But the 545 is the single core variant, the iPad 2 has the dual core variant of the 543, the SGX543MP2. Even if each unit is slower, the iPad 2 has...Well, two. The single core may be faster one on one, but the iPad 2's would still end up faster by a mile. And its not like SLI or Crossfire, the performance scaling is seamless. See here, the 545 isn't on there but you can extrapolate.
Not sweating, I was explaining a point.
As for 32 cores, please note that in desktop computing, the trend of offering processors with more and more cores has decelerated. The same may well happen with mobile devices. Also, please note that we don't need 32 cores to do a billion polygons.
Just checking : )
i agree, which is why i said -if- cores continue. and no cpu cores do not maketh the poly count
or if any of it continues