Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
What Apple has 'traditionally done' applies to 2 devices, both of which needed SoC's that were a step above what they could do in phones at the time and that were, frankly, big compromises. Remember the Retina iPad is... I think I'm right here... the only new version of an iOS device to significantly increase in size and weight over the previous generation. The A5X in particular was the ARM equivalent of the trick some laptop vendors did a few years back of putting desktop parts into 'gaming' laptops as it was the only way to get the necessary power.

We've now reached a point where those compromises don't necessarily have to be made. Actually I'll go further than that, I don't think they *can* be made in the iPad Mini as there just isn't room in the chassis. Everything we've seen about the A7 suggests that it is more than capable on the graphics side with the only slight question mark being over the memory bandwidth. As far as CPU is concerned an A7 would actually be significantly more powerful than the A6X. It would also be far more power efficient, generate less heat and be a considerably smaller package (and, likely, pretty much the same cost if not a little cheaper).

And here's the kicker for me: it makes sense from a marketing PoV. The iPad gets an A7X and a push towards content creation / 'pro' markets hopefully complete with apps. The retina mini gets the A7 (maybe with slight tweaks such as clock speed) and benefits from the 64 bit message Apple will be pushing hard while fitting into a nice neat hole of 'general purpose use / consumption' and selling on price. In other words the same basic strategy they currently use with the Macbook Air and Macbook Pro. It's a message that's simple, is proven to work very well and means that the vast majority of marketing benefits both products.
Great prediction. Wouldn't be surprised either way.
 
The A7's GPU and the A6X's GPU have the same amount of GFLOPs, it will probably perform a little better than the A6X though.

Perhaps Apple will use the A7X SoC in the mini, but rename it to something else like "A7M" and give the A7X in the larger iPad a few hundred MHz clock speed boost for the CPU, to differentiate.

Hmm, but doesn't the X denote for a more powerful graphics unit. They can still use the A7 for the mini and use the A7X for the bigger ipad with perhaps more memory and slightly higher clock speed for the 9.7" ipad i.e 2GB Ram and 1.5Ghz clock speed. The graphics card for the mini will be the same as the one from 5S, while the bigger one could be the G6630. I doubt they would rename it.
 
Hmm, but doesn't the X denote for a more powerful graphics unit. They can still use the A7 for the mini and use the A7X for the bigger ipad with perhaps more memory and slightly higher clock speed for the 9.7" ipad i.e 2GB Ram and 1.5Ghz clock speed. The graphics card for the mini will be the same as the one from 5S, while the bigger one could be the G6630. I doubt they would rename it.
Well the question is...

Is Apple willing to release two iPads (despite the difference in screen size) with the same resolution, but one with twice the graphics power of the other? It's just more work for developers, but perhaps the mini isn't able to handle the power consumption and/or heat output of the A7X.

I can see the mini getting the A7 if it keeps the same 1024x768 resolution... I'm just not sure.
 
Well the question is...

Is Apple willing to release two iPads (despite the difference in screen size) with the same resolution, but one with twice the graphics power of the other? It's just more work for developers, but perhaps the mini isn't able to handle the power consumption and/or heat output of the A7X.

I can see the mini getting the A7 if it keeps the same 1024x768 resolution... I'm just not sure.

I doubt if the A7 is coming in, they will stick to 1024 x 768. Its going Retina, otherwise there is no purpose of switching to A7. The A6 is more than sufficient if no-retina.

The bigger iPad will be the more powerful unit, capable of better graphics performance, more design oriented. Something similar to the lines of the how the Samsung Note 10.1 2014 edition vs Note 8.0. With that in mind, designers can optimize based on A7 rather than A7X. A7X specfic apps will just b more graphic centric. Otherwise its just a simple case of working on apps that are capable in 5S and Mini.

In the bigger picture, I do have a feeling that we might see an A6 powered non-retina ipad mini taking over the cheaper mini spot. Its a feeling based on the assumption that the apple TV, ipod touch and the mini will all move over to the A6, ceasing the end of the A5 line.
 
I doubt if the A7 is coming in, they will stick to 1024 x 768. Its going Retina, otherwise there is no purpose of switching to A7. The A6 is more than sufficient if no-retina.
The purpose would be to give it 64-bit hardware. Also the resolution of the iPhone 5S is near that of the mini's resolution, so it would put it on equal footing. If it does go Retina, and it only uses the A7, then it won't be able to run games at native resolution with the same quality of the 5S.
The bigger iPad will be the more powerful unit, capable of better graphics performance, more design oriented. Something similar to the lines of the how the Samsung Note 10.1 2014 edition vs Note 8.0.
However the 5S packs the same amount of CPU and GPU power as the mini would supposedly have with the A7. So perhaps the mini could handle the A7X.
With that in mind, designers can optimize based on A7 rather than A7X. A7X specfic apps will just b more graphic centric. Otherwise its just a simple case of working on apps that are capable in 5S and Mini.
I know, but that likely means we'll get good quality graphics in games on the A7, but little improvement on the A7X because it will be a smaller percent of the market.
In the bigger picture, I do have a feeling that we might see an A6 powered non-retina ipad mini taking over the cheaper mini spot. Its a feeling based on the assumption that the apple TV, ipod touch and the mini will all move over to the A6, ceasing the end of the A5 line.
Personally I'd prefer that Apple only offer Retina devices from now on. I'd love to see a new Apple TV (as I'm in the market for one soon.)
 
What would be the reasoning for Apple giving the iPad mini 2 their top of the line processor, wouldn't A6X be the logical next step? It would surely be able to drive the rumored retina display, and the mini would still be cheaper and a "step down" from the full size iPad.
The reasoning is that by sometime between 10:00am and 11:00am tomorrow morning (California time) the A7 probably won't be Apple's top of the line processor, the A7X in the full size iPad will be the new top of the line.

I think Apple will put the A7 in the Mini to push 64 bit adoption as quickly as possible and differentiate the full size iPad by giving it an A7X and a 50% chance of 2GB of memory vs the 1GB in the Mini.

I sometimes think that people worry a bit too much about how Apple can differentiate the Mini and the 9.7" iPad range, the different screen sizes, physical dimensions and weight are pretty significant differentiators for most people. I'm really not sure that the average non-Macrumors-reading consumer goes into an Apple store and things "Hmmm, this Mini has an A6X in it vs the A7X in the bigger unit. If only the Mini had an A7 in it I'd have bought it instead of the bigger one.". The thinking is probably much more "Wow, this is really light but I want a bigger screen, but do I want a bigger screen enough to make it worth me spending a bit more money and it not being quite so portable?".
 
I don't think that an A7 chip on a supposed retina iPad Mini makes much sense, but if that is indeed the case, I think we will have the same situation we had with the ipad 3/ipad 4. That is, an update in 6 months because the original release will lag everywhere.

The ipad 4 is suffering from the lack of iOS optimisation for ipad and even the iphone 5s has choppy animations here and there, so unless Apple release an incredible 7.1 update, the A7 on a significantly larger screen will most certainly be a bloody nightmare.

Also because of the aspect ratio of the screen. If the retina iPad Mini will have the same resolution of the larger one in order not to create a nightmare for developers in terms of app scaling, then that means it will have a higher pixel density. That is unless, of course some specific screens were created with the retina resolution and the same ppi of the full size ipad but smaller (which in a way could explain the rumours for production constrains we've been hearing).

Either way, at this point in iOS 7 development, I just don't see the A7 handling an iPad display efficiently.

What someone said about Apple including the A6X in the new iPad mini does make a lot of sense in terms of keeping it at a lower cost.

However, I'm wondering if all the fuss about a Retina mini is justified. It will be the first generation of an iPad mini with a retina display, so I'm wondering if it won't be the same ipad 3 story over again. They release it, performance is generally poor and then a few months later there's a new one to correct the issue.

I've been trying to sell my 4 but now I'm thinking wether or not I should wait until the new ones are released and then see their performance.
 
I don't think that an A7 chip on a supposed retina iPad Mini makes much sense, but if that is indeed the case, I think we will have the same situation we had with the ipad 3/ipad 4. That is, an update in 6 months because the original release will lag everywhere.

Why? The A7 is (broadly) equivalent to the A6X on GPU power, probably around a third slower on memory bandwidth (so roughly in line with the A5X) and has a significantly quicker CPU. It's likely that an A7 in a mini could also be clocked slightly faster than the 5s version. The mini would effectively have roughly the same level of performance as the current iPad and real world would likely feel faster.

Really starting to think people haven't twigged just how powerful the A7 is. Remember the A5X and A6X were both compromises to get sufficient graphical oomph to handle a retina panel, they made almost no improvements to the CPU over their regular counterparts. The A7 CPU in the 5S is quite capable of going toe-to-toe with Intel's upcoming Bay Trail Atom processors (and BT has roughly the same memory bandwidth than the A6X). The GPU goes back and forth depending on the test but look at fill speeds on GLBenchmark and you get:

iPhone 4s - 770
iPad 3 - 1747
iPhone 5 - 1675
iPad 4 - 2059
iPhone 5s - 3,304.5

One more useful benchmark is the Egypt HD test running offscreen (in other words not limited to the refresh rate of the display) at 1080p.

iPhone 4s - 11
iPad 3 - 25
iPhone 5 - 30
iPad 4 - 47
iPhone 5s - 57

The A7 is, in short, a bit of a monster and a true generational leap in performance. It really shouldn't have any problems running the iPad Mini with a Retina display. Frankly, if anything, it's rather wasted in the iPhone! You're effectively looking at a SoC that's months ahead of any equivalent unit from other manufacturers and where it does get beaten in benchmarks it tends to be when the playing field is skewed (e.g. dual core versus quad core or FAR higher clocks).
 
Maybe there is no A7X.

As myself and others point out. The A7 itself should have enough graphical prowess to meet or exceed A6X gpu performance but also blow its cpu performance out of the water.

A7 (non-X) could be used in both iPad 5 and iPad Mini Retina. The reduced power if A7 might be what allows them to thin the iPad 5.
 
Why? The A7 is (broadly) equivalent to the A6X on GPU power, probably around a third slower on memory bandwidth (so roughly in line with the A5X) and has a significantly quicker CPU. It's likely that an A7 in a mini could also be clocked slightly faster than the 5s version. The mini would effectively have roughly the same level of performance as the current iPad and real world would likely feel faster.

Really starting to think people haven't twigged just how powerful the A7 is. Remember the A5X and A6X were both compromises to get sufficient graphical oomph to handle a retina panel, they made almost no improvements to the CPU over their regular counterparts. The A7 CPU in the 5S is quite capable of going toe-to-toe with Intel's upcoming Bay Trail Atom processors. The GPU goes back and forth depending on the test but look at fill speeds on GLBenchmark and you get:

iPhone 4s - 770
iPad 3 - 1747
iPhone 5 - 1675
iPad 4 - 2059
iPhone 5s - 3,304.5

One more benchmark is the Egypt HD test running offscreen (in other words not limited to the refresh rate of the display) at 1080p.

iPhone 4s - 11
iPad 3 - 25
iPhone 5 - 30
iPad 4 - 47
iPhone 5s - 57

The A7 is, in short, a bit of a monster and a true generational leap in performance. It really shouldn't have any problems running the iPad Mini with a Retina display. Frankly, if anything, it's rather wasted in the iPhone! You're effectively looking at a SoC that's months ahead of any equivalent unit from other manufacturers and where it does get beaten in benchmarks it tends to be when the playing field is skewed (e.g. dual core versus quad core or FAR higher clocks).

Thanks for the detailed info. :) I'm just a bit reluctant. The most I'm managing to get for my 4 is 250£. Do you think that's a good price? I think it's a bit too low. It's a 16Gb WiFi only and it's in pristine condition. I would think at least 285£ would be a fair price for it.
 
I think the Mini 2 will get the A7 while the iPad 5 gets a A7X to be able to boast a big increase in performance over the earlier model. I base this both on their marketing strategies and that they want to put their top tier components in their more expensive products. The Mini is dirt cheap so I actually believe you can't expect it to have the same hardware as its bigger sibling.

Hopefully I can get a good selling price for my iPad 4 with 3G and 32 GB when the iPad 5 arrives at my house, otherwise i'll just give it to my kid sister who is still using my old iPad 1.
 
Thanks for the detailed info. :) I'm just a bit reluctant. The most I'm managing to get for my 4 is 250£. Do you think that's a good price? I think it's a bit too low. It's a 16Gb WiFi only and it's in pristine condition. I would think at least 285£ would be a fair price for it.

Ouch. Looking at CEX they'd give you £273 in trade so somewhere around £285 wouldn't seem unreasonable to me. That being said loosing a third in a year for a device with a fairly limited shelf life isn't *that* bad, especially if the next gen is a significant change. Sorry, this isn't as helpful a reply as I'd like it to be, guess it depends how much you value the extra £35 versus having cash in your pocket (presumably to feed your Apple Acquisition Disorder :D)

No problem on the info either, wish more gadget sites would do proper reviews. Seems a shame that Anandtech is often the only one that really does a deep dive into this stuff.
 
Ouch. Looking at CEX they'd give you £273 in trade so somewhere around £285 wouldn't seem unreasonable to me. That being said loosing a third in a year for a device with a fairly limited shelf life isn't *that* bad, especially if the next gen is a significant change. Sorry, this isn't as helpful a reply as I'd like it to be, guess it depends how much you value the extra £35 versus having cash in your pocket (presumably to feed your Apple Acquisition Disorder :D)

No problem on the info either, wish more gadget sites would do proper reviews. Seems a shame that Anandtech is often the only one that really does a deep dive into this stuff.

Yeah, I popped by CEX the other day. Their cash value trade in is a bit offensive to be honest, but I found another small store that was willing to five me 250£ in cash. I didn't try to negotiate and just enquired, so I think I'll pop by and try to negotiate a few more pounds! :)
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.