Next-Gen iPhone & iPhone Nano Speculation

[ltcommander.data]

http://www.theverge.com/2012/3/6/2849384/sources-apple-tv-update-ipad-3-ipad-hd-lte-verizon-att

Joshua Topolsky's final prediction is that the iPad 3 will be using the A5X afterall, reportedly dual core with improved GPU.

Curiously he believes the A6 will still see the light this year in the iPhone 5. If the Apple A6 is quad core Cortex A9, it seems strange that the iPad would miss out on it. If it's merely a production timing issue, the iPhone doesn't seem to desperately need a quad core CPU, so why not just stick with the smaller, by that time mature A5X for the iPhone 5? I suspect then that if the iPad 3's A5X is high clock speed Cortex A9 + SGX543MP (2-4) and the Apple A6 is still coming this year in the iPhone 5, then the Apple A6 is not going to be quad core Cortex A9, rather dual core Cortex A15 + same GPU as the A5X (low-risk option) or perhaps Rogue (high-risk option).

 

[Entropy]

What it really does is uncover the true cost of these phones, which are very much hidden when you subsidize them. The average person wouldn't know that an iPhone is $100 more expensive than every other phone out there unless they had to replace it before their contract ended.

Well, around here everyone knows what an iPhone costs, and it is still the dominant smartphone. As far as I'm aware, phones are sold contract-less in most parts of the world, including the worlds largest cell phone markets. The US is an anomaly.

 

[wco81]

There was an article about iPhone losing ground in the PIIGS countries.

Also iPhone lost share recently in Germany.

 

[Entropy]

There was an article about iPhone losing ground in the PIIGS countries.

Also iPhone lost share recently in Germany.

Dependent on how you want to spin things, you can pick and choose specific markets, and specific quarters. Apple is consistent enough that you can always find quarters and markets to support any headline. Longer term it seems clear that Apple want to maintain a strong grip on the parts of the market which is profitable (for them, and incidentally, for software developers), and care less about the low end of the market. Logically, this should lead to Apple loosing total market share over time. If this matters in the slightest to them or developers for their platform is more debatable, as long as they dominate the market segment that is actually prepared to pay for what they consume.

 

[Deleted member 13524]

http://www.engadget.com/2012/03/07/ipad-rumor-tactile-display/

Tactile display, which some of us thought it would end in Wii U's tablet controller, might be coming to iPad 3.

 

[wco81]

Dependent on how you want to spin things, you can pick and choose specific markets, and specific quarters. Apple is consistent enough that you can always find quarters and markets to support any headline. Longer term it seems clear that Apple want to maintain a strong grip on the parts of the market which is profitable (for them, and incidentally, for software developers), and care less about the low end of the market. Logically, this should lead to Apple loosing total market share over time. If this matters in the slightest to them or developers for their platform is more debatable, as long as they dominate the market segment that is actually prepared to pay for what they consume.

Well they're happy with the profits but I'm sure they care about Android selling so well, especially when they think Android infringes on their IP. At least that's what Steve Jobs thought and reportedly turned down an offer from Eric Schmidt to make amends.

But now, there's a story out that Apple proposed selling Motorola and Samsung, with which they're involved in several lawsuits, patent licensing at $15 a handset.

 

[Entropy]

Well they're happy with the profits but I'm sure they care about Android selling so well, especially when they think Android infringes on their IP. At least that's what Steve Jobs thought and reportedly turned down an offer from Eric Schmidt to make amends.

I think part of Apple/Jobs' stance there is because of history repeating itself. Microsoft working with early Apple Lisa/Mac to produce software for the new machine taking key ideas and running off to produce a competing platform with (in the eye of the public) similar functionality. Eric Schmidt on Apples' board of directors, taking advance info on their new products home to his own company to produce a competing platform with (in the eye of the public) similar functionality.

By gods, it must have been galling. I can understand that it was unforgivable.

In other news, I hope we will get some solid info regarding the SoC in the new iPad tonight.

 

[Arwin]

http://www.engadget.com/2012/03/07/ipad-rumor-tactile-display/

Tactile display, which some of us thought it would end in Wii U's tablet controller, might be coming to iPad 3.

If that happens, I would expect the resolution increase to not happen, and vice versa. But we'll see soon enough.

 

[Lazy8s]

543MP4 it is!

 

[benjiro]

543MP4 it is!

Yep. But looks like a bit of marketing:

Comparing to a Tegra 3, Apple says the A5 is "twice as fast" and the A5X offers "four times the​ performance."

Last time i checked, the Tegra3 was almost as fast as the A5. Or was able to produced the same or better image quality ( with there Tegra Optimized Games ).

Now, twice as fast as that, must be a 543MP4. Looks like Apple will hold the performance crown on GPU for the next year.

 

[Lazy8s]

They may have kept the same clock speeds, including the 250MHz on the 543MP4, as I expected.

Seems to still be dual core A9.

So, A5X for iPad 3. The A6, then, would be for either the next iPhone or simply next year…

Seems to me Apple nailed it perfectly: staying out of the futile CPU and MHz race yet pushing the GPU to more cores for the iPad 3's 4x resolution display.

 

[Deleted member 13524]

543MP4 it is!

And no mention on the CPU, so I'd guess dual-core Cortex A9.
Also, LTE.



Lots of comparisons with products from the competition this time.. First mentioning HP PC sales, then getting a big shot of Samsung Galaxy Tab 10.1 in there, calling it "just a big phone" (yeah, and the iPad is what?), then mentioning Tegra 3's performance..

Why would they feel the need to do that? Are they scared?

Oh, and no Siri. lol?

 

[Helmore]

Yep. But looks like a bit of marketing:

Last time i checked, the Tegra3 was almost as fast as the A5. Or was able to produced the same or better image quality ( with there Tegra Optimized Games ).

Now, twice as fast as that, must be a 543MP4. Looks like Apple will hold the performance crown on GPU for the next year.

Last time I checked, the Apple A5 in the iPad 2 was roughly twice as fast as the Tegra 3 in the ASUS Transformer Prime in GLBenchmark PRO on 720p. That's just one benchmark though, but that's all that Apple needs to justify their claim.

 

[Ailuros]

Not really twice as fast. iPad2 is in Egypt 720p at over 10k frames and the Transformer TF201 at over 7.7k frames. In the PRO test yes, but that's not really a field to concentrate on IMO for OGL_ES2.0 GPUs.

 

[Lazy8s]

1GB of RAM apparently

 

[Helmore]

Not really twice as fast. iPad2 is in Egypt 720p at over 10k frames and the Transformer TF201 at over 7.7k frames. In the PRO test yes, but that's not really a field to concentrate on IMO for OGL_ES2.0 GPUs.

I agree with you, but Apple only really needs one benchmark that supports their claims to not be considered a liar.

 

[ltcommander.data]

then getting a big shot of Samsung Galaxy Tab 10.1 in there, calling it "just a big phone" (yeah, and the iPad is what?)

http://live.theverge.com/Event/Live_from_Apples_iPad_3_event_in_San_Francisco?Page=2

1:20 pm Ouch — showing Twitter for Android tablets. "It kinda looks like a blown up smartphone app... that's because that's what it is!"

I'm not sure if this quote from the Verge's liveblog refers to what you are talking about, but it seems Apple was talking about Android tablet apps often being blown up smartphone apps rather than the tablet device itself. I don't know the latest comparison figure, but I would think pointing out the App Store's higher number of tablet optimized apps is a true assessment of the situation and a fair usability benefit to iPad customers.

On another note I wonder why they are going back to calling them iSight cameras instead of FaceTime/FaceTime HD cameras?

 

[Deleted member 13524]

Last time I checked, the Apple A5 in the iPad 2 was roughly twice as fast as the Tegra 3 in the ASUS Transformer Prime in GLBenchmark PRO on 720p. That's just one benchmark though, but that's all that Apple needs to justify their claim.

Egypt is only ~30% faster than Transformer Prime, so I'll call the 4x faster than Tegra 3 (in any possible perspective) quite a bit of BS.

Furthermore, I'd guess the Tegra 3 versions using DDR3L 1500MHz memory may get quite a bit faster 3D performance than the current 1066MHz LPDDR2 tablets. Tegra 3's dealing with a single 32bit channel, after all.

The "more memory than X360" points to 1GB RAM.

 

[ltcommander.data]

Egypt is only ~30% faster than Transformer Prime, so I'll call the 4x faster than Tegra 3 (in any possible perspective) quite a bit of BS.

Furthermore, I'd guess the Tegra 3 versions using DDR3L 1500MHz memory may get quite a bit faster 3D performance than the current 1066MHz LPDDR2 tablets. Tegra 3's dealing with a single 32bit channel, after all.

The "more memory than X360" points to 1GB RAM.

We don't know the GPU clock speed though. 4x could be accurate if the clock speed has also been increased relative to the iPad 2.

Is DDR3L inherently faster than LPDDR2? Otherwise would 1x32-bit 1500MHz DDR3L really be faster than 2x32-bit 1066MHZ LPDDR2?

 

[Deleted member 13524]

We don't know the GPU clock speed though. 4x could be accurate if the clock speed has also been increased relative to the iPad 2.

There were two statements regarding 3D performance:
1 - A5 is 2x faster than Tegra 3 (a factual lie)
2 - A5X is therefore 4x faster than Tegra 3, because it's 2x faster than A5.

Is DDR3L inherently faster than LPDDR2? Otherwise would 1x32-bit 1500MHz DDR3L really be faster than 2x32-bit 1066MHZ LPDDR2?

I don't know if it's inherently faster but I doubt it's inherently slower.
Nonetheless, I don't understand your question.
The Transformer Prime (TF201), the only Tegra 3 implementation in the market so far, is using a single-channel 32-bit LPDDR2 @ 1066MHz. Asus TF301 will use 32bit single-channel DDR3L @ 1500MHz.
So it may get a bit faster on GPU-intensive applications.