Next-Gen iPhone & iPhone Nano Speculation

While the potential to spread clocks among more cores and use a lower voltage to save power does exist for quad core, the proportion of the workload that actually takes advantage of that is and probably will continue to be minimal. Mobile OSes are fairly good about minimizing work done on off-screen tasks (browser tabs included), and, from the admittedly little power profiling I have seen, webpages require a burst mode at peak workload from multiple CPU cores even less frequently than I already imagined.

I still believe more cores can be a great design trade-off because of the extra flexibility afforded, yet those extra cores should be low power CPUs matched closely to the typical usage scenarios (background tasks waking up for a quick data sync, etc.)
 
Here is something new to discuss:

http://www.patentlyapple.com/patent...cture-will-it-take-ios-to-the-next-level.html



Its a interesting concept, that's for sure.

I would think it would be a pretty big jump if they went to lengths to trademark it. Trademarking means they intend to refer to it when discussing things with customers, and most customers don't give a damn about CPU details.

edit: http://www.theverge.com/2012/2/8/2785486/ipad-3-back-photo-appears-a6-retina-display

A6 staying dual core CPU, beefier GPU?
 
Last edited by a moderator:
I would think it would be a pretty big jump if they went to lengths to trademark it. Trademarking means they intend to refer to it when discussing things with customers, and most customers don't give a damn about CPU details.

edit: http://www.theverge.com/2012/2/8/2785486/ipad-3-back-photo-appears-a6-retina-display

A6 staying dual core CPU, beefier GPU?

would make sense if they are still 40nm . Get one more gen out of it in the ipad before they give the quadcore to the cell phone later in the year ?
 
would make sense if they are still 40nm . Get one more gen out of it in the ipad before they give the quadcore to the cell phone later in the year ?

The problem with that theory is, that they seem to have the habit of first trying there new stuff in the iPad, and then releasing it in the Cell Phone. What makes sense... They can build up software, have less worries with battery ( bigger battery ), etc.

If they stick with a Dual Core, it means two possibilities: A9 or A15 based CPU.

A faster A9 dual core, is not going to cut it. Last generation is Dual 1Ghz, while Tegra3 is already Quad 1.3Ghz ( with 1.4Ghz Single Core boost ). Assuming they stay with same manufacturing process. If they go with 28nm, then there really is no reason why they can not go Quad Core?

Unless they want to increase the Speed to extreme high on the same manufacturing process, but we all know that higher speed = more battery drain ( its like trying to overclock a CPU on a manufacturing process, there is a limit on how fast you can clock a CPU without exponentially increasing the power consumption & heat output ).

Or a A15, forget about seeing a A15 on same manufacturing process. That is just silly to design a new architecture in there A6 SOC, and stick to same manufacturing process. It will limit the speed.

And Apple Waiting another year while staying at same manufacturing process... And why increase the battery design then? Only for the screen...

Then its not going to be a iPad3, but a iPad2s.

Don't know, something smell wrong with this picture. Especially how little information there is. With those Quad Core "leaks" you had some visual information, that fit. CPU Number, Quad Core, 1GB memory, ...

But now all we get is a dual picture with bigger battery claim ( Check ), and for the rest not a single piece of evidence to back up the other claims. Hmmmm ...

Lets just say, i'm very skeptical about this claim. To me it does not make sense to stay Dual core, unless they can go A15 28nm.

Simply put, are there mass people going to by a iPad3(2s), with these specs? If people do not see it as a major upgrade, in general people with a iPad2 will stick to there "old" one = less sales.

And you also run the risk, that the iPad3(2s) is "old", we got Tegra3 already out, Tegra3 ( 28nm not far off ), Krait ( 28nm ), and several other designs coming out in the next few months.

And seeing as they mention a beefier GPU, but CPU is just mentioned as Dual core. The only way i see a Dual Core design in the iPad3, while being considered a major upgrade from the iPad2, if the push a A15 28nm in there.
 
The problem with that theory is, that they seem to have the habit of first trying there new stuff in the iPad, and then releasing it in the Cell Phone. What makes sense... They can build up software, have less worries with battery ( bigger battery ), etc.

If they stick with a Dual Core, it means two possibilities: A9 or A15 based CPU.

A faster A9 dual core, is not going to cut it. Last generation is Dual 1Ghz, while Tegra3 is already Quad 1.3Ghz ( with 1.4Ghz Single Core boost ). Assuming they stay with same manufacturing process. If they go with 28nm, then there really is no reason why they can not go Quad Core?

Unless they want to increase the Speed to extreme high on the same manufacturing process, but we all know that higher speed = more battery drain ( its like trying to overclock a CPU on a manufacturing process, there is a limit on how fast you can clock a CPU without exponentially increasing the power consumption & heat output ).

Does it matter ? The second gen ipad only had 512 megs of ram while other tablets had 1 gig

Also remember the iphone kinda got a half assed release with the 4S . So perhaps the reason is that the new cpu is not ready and may not be ready with the ipad 3 .

Ipad 3 with higher res screen + 1 gig of ram + faster cpu and gpu could be more than enough to sell more of them

Or a A15, forget about seeing a A15 on same manufacturing process. That is just silly to design a new architecture in there A6 SOC, and stick to same manufacturing process. It will limit the speed.

And Apple Waiting another year while staying at same manufacturing process... And why increase the battery design then? Only for the screen...

Then its not going to be a iPad3, but a iPad2s.

Don't know, something smell wrong with this picture. Especially how little information there is. With those Quad Core "leaks" you had some visual information, that fit. CPU Number, Quad Core, 1GB memory, ...

But now all we get is a dual picture with bigger battery claim ( Check ), and for the rest not a single piece of evidence to back up the other claims. Hmmmm ...

Lets just say, i'm very skeptical about this claim. To me it does not make sense to stay Dual core, unless they can go A15 28nm.

Simply put, are there mass people going to by a iPad3(2s), with these specs? If people do not see it as a major upgrade, in general people with a iPad2 will stick to there "old" one = less sales.

And you also run the risk, that the iPad3(2s) is "old", we got Tegra3 already out, Tegra3 ( 28nm not far off ), Krait ( 28nm ), and several other designs coming out in the next few months.

And seeing as they mention a beefier GPU, but CPU is just mentioned as Dual core. The only way i see a Dual Core design in the iPad3, while being considered a major upgrade from the iPad2, if the push a A15 28nm in there.

People will buy anything with an a little i in front of it , its been proven many times , the ipad 2 wasn't the highst spec tablet when it was released or today but it still sells the best. The iphone isn't the best speced phone but it still sells extremely well.

It could simply be that the new chips are not ready and this is a stop gap ala the iphone 4s . It will help them remain competetive this year with the quad core devices coming out and get consumers to upgrade
 
People will buy anything with an a little i in front of it , its been proven many times , the ipad 2 wasn't the highst spec tablet when it was released or today but it still sells the best.

Maybe on CPU end it was not extreme high end at that time, but on GPU it really destroyed any competition on the market.

Do not forget, you need to look at it from iPad1 -> iPad2. This was a massive move forward. The CPU doubled it cores, and was faster. The GPU crushed the iPad1's GPU ( and any other competition. Even now, the the competition has finally cached up ).

This is the problem. Do people want to upgrade from a iPad2 -> iPad3, with some small speed upgrades. In other words, a iPad2s? Knowing that in a short while, the technology is massively jumping again.

Something where you spend a lot of money into ( lets face it, Apple is not cheap ), needs to be worth it ( that include being future proof ).

Your not going to spend the full amount on a slight upgraded iPad2s, when you can get discounted / second hand / refurbished iPad2 for way cheaper, that offers you almost the same experience, just 10 or 20% slower ( assuming its just a small upgrade )?

We shall see. Probably a few weeks from now, we will get the official news ( and probably some more leaks also ).
 
Cortex-A15 isn't coming to any end product shipping within the next few months.

Process availability and capacity is almost always lagging a fab's optimistic projections, so any node beyond the current 45nm might be a tight schedule to make depending on when Apple's A6 is introduced.

Apple should stick with relatively low clocked A9s this generation and then move to their custom ARMv7 design for the A7 SoC, assuming they were up to the design challenge. They don't need to move on to Cortex-A15 at all, but it depends on whether their custom design is for a high power core, a low power core, or whether they're designing both.

Graphics performance will definitely be improved, but they jumped so far ahead last time that they can afford a minor upgrade from a clock speed increase and still not lag the market by too much in the coming year/product cycle.

As long as they make sensible advances and improve the user experience through better software, consumers won't care about the specifics of the tech specs.
 
Yeah, I think people are getting ahead of themselves when just discussing hardware specifications here.

The fact remains that iOS require less resources than Android so far and that the user experience is much, much smoother on iOS, even when it is supposedly lagging in terms of raw hardware.

But you can be rest assured that Apple is not resting on its laurels. Even though they do not actively play the specification game, they seem to take their SoCs very serious despite it.

There really still isn't a competitor to the A5 in terms of overall performance in a shipping product today.

So, what would make sense from here on out? Custom ARMv7? Quad-Core Cortex-A9? The graphic performance could easily surpass any other SoC with a SGX543MP4 clocked at the same 250 Mhz.

Sure, we all dream about Cortex-A15 but then the A6 really need to be on 28 nm but everything else seems to indicate it is on Samsung's 32nm process.

It also seems that the battery compartment in the iPad 3 is larger, that coupled with a "Retina" Display alone would make it sell like hot cakes.
 
People will buy anything with an a little i in front of it , its been proven many times , the ipad 2 wasn't the highst spec tablet when it was released or today but it still sells the best. The iphone isn't the best speced phone but it still sells extremely well.
Both iPhone (4S) and iPad (2) had and still have the most powerful GPU in any mobile device. The newest Samsung products (Galaxy Note, Nexus and the new 7" Galaxy Tab) come close, but do not match either iPhone or iPad. iPhone still has the best DPI of any display, and iPad 2 IPS display quality is also one of the best in market. The specs certainly do matter. Apple hardware specs (related to graphics rendering and outputting) are top of the class.
 
benjiro said:
This is the problem. Do people want to upgrade from a iPad2 -> iPad3, with some small speed upgrades. In other words, a iPad2s? Knowing that in a short while, the technology is massively jumping again.
Give me an 4x resolution screen with the same specs and I'm already on the fence. Up the CPU and GPU performance just a little bit and you have a sale.

When I use my old iPad 1 (upgraded to iOS5), it often surprises me how snappy it is. With the iPad 2: no issues at all. The screen is where it's at.
 
Give me an 4x resolution screen with the same specs and I'm already on the fence. Up the CPU and GPU performance just a little bit and you have a sale.
Same here... The thing that needs the upgrade most is the screen resolution. Otherwise the iPad 2 is pretty perfect for what it does the best (it could however be slightly lighter though).
 
Last edited by a moderator:
I've been waiting and waiting, especially after rumors of the higher DPI screen circulated last year.

Meanwhile, I've been downloading/buying and adding to my wish list a bunch of iPad apps.

If I relied only on the roadmap of SOC designs, I'd never buy. But no matter how powerful these SOCs become, they won't reach the raw CPU performance of the Sandy Bridge E i7 in my MacBook Pro. So I'm not going to process RAW files on an iPad any time soon, when doing so pegs the i7.

As far as weight, I guess that's the tradeoff for battery life. At least 90% of the time, my laptop is plugged into the outlet. With tablets, you expect that ratio to be reversed, so you need a beefy battery.

For extended reading in bed, now that e-ink readers are so affordable, you can get both a tablet and a dedicated reader.
 
I've been waiting and waiting, especially after rumors of the higher DPI screen circulated last year.

Meanwhile, I've been downloading/buying and adding to my wish list a bunch of iPad apps.

If I relied only on the roadmap of SOC designs, I'd never buy. But no matter how powerful these SOCs become, they won't reach the raw CPU performance of the Sandy Bridge E i7 in my MacBook Pro. So I'm not going to process RAW files on an iPad any time soon, when doing so pegs the i7.

As far as weight, I guess that's the tradeoff for battery life. At least 90% of the time, my laptop is plugged into the outlet. With tablets, you expect that ratio to be reversed, so you need a beefy battery.

For extended reading in bed, now that e-ink readers are so affordable, you can get both a tablet and a dedicated reader.

Well, there will be some turning point for sure.

What hardware did you process your first RAW image on? Because I can recall doing it on an old eMac PowerPC G4 with 512 MB ram and earlier models.

I could easily see Aperture being ported to iOS and the ability to process RAW images on a quad-core Cortex-A9. Sure, you are not going to batch process a whole bunch of images but using a Retina Display iPad as a field monitor (and preview monitor on set) would be plain awesome.
 
I'm not going to lug around an iPad out in the field. Maybe take it on trips in addition to the laptop but won't take it outside, though I'm seeing more and more iPads on subways.

Saw two women use an iPad to take pictures inside the Cordoba Mezquita the other day. One holding it up and the other tapping the shutter activation button.

Apple has very unfavorable pricing for storage, the $100 bumps are not a good value. Most I'd get is 32 GB (and sometimes people predict double the capacity at same price points but why would they bother when their current SKUs sell well?).
 
Apple has very unfavorable pricing for storage, the $100 bumps are not a good value. Most I'd get is 32 GB (and sometimes people predict double the capacity at same price points but why would they bother when their current SKUs sell well?).

True. Never understood the reasoning behind having fixed storage, and no SD support. Going from a 16GB->32GB model, has a 100$ price increase, while a 16GB SD card is ... what 20 or 30$.

While the internal storage can be faster then any SD card, a lot of the data you end up in there ( like Music, Video, etc ), those thing do not require a fast access / transfer time. A slow 4 to 5MB/s is about all you need for music / video playing, and its this what takes the most space.

Its silly that people need a "iPad camera kit" to easily access there pictures.
 
True. Never understood the reasoning behind having fixed storage, and no SD support. Going from a 16GB->32GB model, has a 100$ price increase, while a 16GB SD card is ... what 20 or 30$..

Which part of getting an extra $100 for something that might only otherwise sell for $30 don't you understand :)

I'd wager they sell at least as many 32gb ipads as 16gb.

However, as far as I'm concerned, the reason is to do with ecosystem control. It's far from trivial for the average person to get non-itunes/ripped/downloaded content into/out of an ipad or iphone compared to other platforms. While to users that may seem restrictive, to content providers its brilliant. It means the vast majority of paid content on ios is actually paid for, hence content providers see ios as more secure platform and are more aminable to provide for it. The fact that to get software onto an iphone requires it go thru apples approval process also means the chances of there being software available to circumvent the above is much reduced (I'm ignoring jailbreaking as its pretty irrelevant in real terms).

It is not by accident that there is no USB, no external storage, and no file transfer via bluetooth on ios.

I am a bit surprised that apple allows the likes of ( errr I think its called AVplayerHD, don't have my ipad to hand), that allows playing of virtually any movie format and more importantly supports for wifi (non-itunes) transfer of media.
 
Last edited by a moderator:
I'm not going to lug around an iPad out in the field. Maybe take it on trips in addition to the laptop but won't take it outside, though I'm seeing more and more iPads on subways.

Saw two women use an iPad to take pictures inside the Cordoba Mezquita the other day. One holding it up and the other tapping the shutter activation button.

Apple has very unfavorable pricing for storage, the $100 bumps are not a good value. Most I'd get is 32 GB (and sometimes people predict double the capacity at same price points but why would they bother when their current SKUs sell well?).

I tell you, photographers and videographers alike are craving for this one feature.

To be able to use the iPad (especially if it gets a Retina Display) to use as a field monitor / preview monitor.

Honestly, the price brackets they have chosen works. People buy it at those prices as you say yourself.
 
I've been waiting and waiting, especially after rumors of the higher DPI screen circulated last year.

Meanwhile, I've been downloading/buying and adding to my wish list a bunch of iPad apps.

If I relied only on the roadmap of SOC designs, I'd never buy. But no matter how powerful these SOCs become, they won't reach the raw CPU performance of the Sandy Bridge E i7 in my MacBook Pro. So I'm not going to process RAW files on an iPad any time soon, when doing so pegs the i7.
If you want to do professional image processing or painting on a tablet, the Samsung Slate is a better fit for that purpose than iPad 2. The Slate has a 128 GB SSD, 4 GB of memory, Sandy Bridge based i5 CPU (dual core with HT = 4 threads), 11.6" 1366x768 IPS screen and it runs all x86 programs (Windows 7) such as Photoshop and After Effects. It also has 1024 level pressure sensitive Wacom digitizer and a (full sized) Wacom pen (perfect for portable painting). USB port allows you to plug in your DSLR or video camera and copy files from them, and your external USB drives if you need to edit long videos (128 GB SSD lasts only for so long). And if you need to connect your photo printer, you can directly connect it as well (the device has full Windows 7, so you can install all Win 7 device drivers to it). It also connects to Windows networks, so you can access network drives/printers easily at home/office (and move data easily). It also has bluetooth connectivity, and comes with a (ultra lightweight) wireless keyboard. So you can transform it to a "desktop computer on the go" if you need to do some serious work.

It weights slightly more than iPad 2 and has worse battery life, so it isn't as good for consumer usage patterns (light web browsing, pdf reading, email reading, etc). But for professional usage, it's currently the best touchscreen computer available. Compared to Wacom Cintiq 12" drawing board, the Samsung tablet price (1200$) isn't that bad at all. The board costs 999$, and the Samsung Slate has a full Sandy Bridge i5 based computer + SSD added for just 200$ extra. Of course it's a niche product, but if you fall to that niche, the product is actually really good... Still it's a hard device to recommend, as the Ivy Bridge based Win8 version will likely be a much more polished all around device (more performance, better battery life, better user interface).
 
Back
Top