blackcrayon

mainyehc said:

Wait, maybe I’m missing something here, so please bear with me:

AppleInsider said:

[…]

Before we make the Apple II sound like a machine you would want to buy today, you wouldn't. For all its firsts, it had a major deficiency in that only supported uppercase letters. If you bought an Apple II in 1977, you could only type on it in capitals. It wasn't until 1983 and the Apple IIe that it shipped with the ability to show lowercase too.

[…]

But then suddenly it was the 1990s, and still the Apple II was selling. By this point, it had gone through very many variations, but the final one was the Apple IIe. It ceased production in November 1993.

How could the Apple IIe have been simultaneously launched in 1983 *and* the “final” model? The final one to be actually in production, standing alone, after a 10-year-long run? You’re meaning to tell me Apple discontinued the more capable Apple IIGS even before the IIe, and not the entire remaining lineup at the same time? I mean, I know Apple is no stranger to selling really old hardware for years without updating it – the Macintosh Plus, the iPod Classic, the iPod Touch, the Mac Mini and the infamous 2013 Mac Pro come to mind –, but jeez, ten years?

Edit: ooooh, I see. That would’ve been indeed the Apple IIe Platinum. And judging from its specs, it wasn’t that different from the other IIe models. Still extremely weird, IMHO.

Yep, I often wonder what it would've been like if Apple simply discontinued the IIe in 1987 or so, and all current and future orders for IIe's were simply filled by IIgs machines.  Since the GS is highly compatible with the IIe, almost all software and hardware works.  I guess the obvious answer is cost- Making a IIe in 1989, which wasn't much different than it was in 1983, must've only cost Apple "pennies" compared to making the relatively complex IIgs.

crowley said:

AppleInsider said:

It's the Apple II that made the company, kept it afloat, and even made it a cult success -- but it was also the Apple II that Steve Jobs tried so hard to kill off with the Mac. It's the little machine that could, did, and for its fans, still does.

Still does?  Not seeing much evidence of that in the article.  Is anyone still using an Apple II for anything other than nostalgia?

I guess "nostalgia" can be defined a bit, but plenty of us use Apple II's (albeit later models than the absolute *original* Apple II) for fun. There are even new peripherals that get developed for them - video cards, storage devices, and it's still a fun challenge making a 35-40 year old computer do things it was never intended to.

bb-15 said:

blackcrayon said:

Onscreen tests are capped at the refresh rate of the display which is usually 60Hz... Which is why you won't see any scores above 60 (check over at Anandtech where they test a lot of devices).  Eventually the GPUs get so good at the older GL benchmarks that just about everything gets around 60 onscreen unless there is a drastically different native resolution of the screen.  And then they have to update the benchmarks to make it worthwhile to even run onscreen tests.

Offscreen tests are done at 1080p, but there is no display to hold back the refresh rate, so it's unlimited.  It's interesting to compare the two, because you get situations where "hey the Galaxy S6 GPU is better than the iPhone 6's GPU, it kills it in offscreen tests!" but then "Yeah but the Galaxy S6 GPU still can't handle its high native screen resolution so XYZ looks choppy!"...

I think we'll probably have the usual effect of the Pro only looking slightly faster at first, but the difference becoming more apparent as iOS grows e.g. iOS 10.

Thank you for the explanation. So, with the gfxbench onscreen GPU tests done by Ars Technica, the T-Rex HD results of the iPad Pros could be ignored since the Pro devices are bumping up against the 60 HZ limit of the displays and are almost at 60 fps with that test.
- But Ars also did a Manhattan HD onscreen test. And the scores for that did not bump up against the 60 HZ / 60 fps limit.
- 9.7 iPad Pro 35.3 fps
- 9.7 inch Air 2 28.1 fps
Back to my point. The graphic performance of the new iPad and the Air 2, on their screens (which is currently what most users will see), does not seem to be that different.

I don't think I'm disagreeing with you as much as you seem to think  But yes, the larger difference will be seen in CPU speed, where the Pro is significantly faster for more important single core tasks.  In any case, the Anandtech review (or maybe the full Ars one) should have a whole lot more comparisons that include more CPU tests and things like throttling, etc.   I'm getting the Pro because I want the Pencil support and any increases in speed-  but I will use them.  The Air 2 is still a great value and probably the second best 10" tablet on the market.

AI2xxx said:

blackcrayon said:

Offscreen tests are done at 1080p, but there is no display to hold back the refresh rate, so it's unlimited.  It's interesting to compare the two, because you get situations where "hey the Galaxy S6 GPU is better than the iPhone 6's GPU, it kills it in offscreen tests!" but then "Yeah but the Galaxy S6 GPU still can't handle its high native screen resolution so XYZ looks choppy!"...

I think we'll probably have the usual effect of the Pro only looking slightly faster at first, but the difference becoming more apparent as iOS grows e.g. iOS 10.

Offscreen is always the most relevant in defining an SoC's true performance. Content and applications don't need to be running at the device's native resolution. Additionally, GPU compute does not have to depend on the resolution of the screen. Onscreen comparisons had been popular in a time when it would take a high end phone or tablet just to run the UI and basic applications properly. As with the PC world, offscreen or equal resolution comparisons are always used.  

I realize that, in fact that's why I gave the example I gave.  It *is* important that a GPU be able to handle the resolution that the screen forces it to though.  If I put the most powerful mobile GPU in the market paired with a 4K screen, it's not going to be enough to drive it particularly well, so things will have to be either low frame rate or run at a quality level/resolution that wastes the higher quality screen.  So I'm sticking with both onscreen and offscreen tests as being important benchmarks.
Onscreen comparisons are important with mobile devices because they have fixed screen resolutions and mobile GPUs are still very weak compared to what we're used to on desktops.

bb-15 said:

blackcrayon said:

Which graphs are you looking at?

The Geekbench 3 single core shows the new Pro is ~70% faster than the Air 2. (And ~15% faster in multicore).
And the onscreen GPU test is capped at 60 fps.  Which as you said won't matter for most apps.  But the fact that offscreen tests show the Pro is ~50% faster than the Air 2 means it has more breathing room for more graphically intense software.

The Air 2 should be fine for a while and I agree it's great for most users/tasks, but the at least 50% improvement in CPU and GPU on the Pro shouldn't be unnoticeable.

- The onscreen gfxbench GPU test scores I wrote about were straight from the Ars Technica graph. I saw nothing about the gfxbench test being capped at 60 fps.
About this Cunningham said; "Onscreen performance, which renders scenes at the screens' native resolutions, is about the same (as the big iPad Pro)" But it is also not that much faster than the Air 2 as the scores show.
- As for the off screen gfxbenchGPU test (which is always done at 1080p) the scores are way above 60 fps which again does not show any cap. The higher off screen score shows the power of the GPU in the new iPad which could eventually have better performance with higher resolution games though it's hard to say when those will be released.
* As for what would be noticeable right now; software that is CPU intensive which uses single core will clearly be faster on the new iPad Pro. But current apps that are multicore intensive or available GPU intensive apps used on each device's native resolution are not going to show much of a difference imo.

Onscreen tests are capped at the refresh rate of the display which is usually 60Hz... Which is why you won't see any scores above 60 (check over at Anandtech where they test a lot of devices).  Eventually the GPUs get so good at the older GL benchmarks that just about everything gets around 60 onscreen unless there is a drastically different native resolution of the screen.  And then they have to update the benchmarks to make it worthwhile to even run onscreen tests.

Offscreen tests are done at 1080p, but there is no display to hold back the refresh rate, so it's unlimited.  It's interesting to compare the two, because you get situations where "hey the Galaxy S6 GPU is better than the iPhone 6's GPU, it kills it in offscreen tests!" but then "Yeah but the Galaxy S6 GPU still can't handle its high native screen resolution so XYZ looks choppy!"...

I think we'll probably have the usual effect of the Pro only looking slightly faster at first, but the difference becoming more apparent as iOS grows e.g. iOS 10.