Broadcomm has the Adreno,
Qualcomm has Adreno. (Readon anagram since this is mobile graphics spun out of ATI back before they were acquired by AMD ).
Imagination licenses the PowerVR to whoever wants to use it,
Including Intel.
"... While Intel uses its Gen graphics in Bay Trail, the GPU in both SoCs is still from IMG. Merrifield features the PowerVR Series 6 G6400, while Moorefield uses the G6430. Both are four cluster designs, the latter is just optimized for higher performance. This is roughly the same GPU configuration Apple uses in the iPhone 5s/iPad Air, but at somewhat higher frequencies from what I can tell (and of course, built on Intel's 22nm process and not Samsung's 28nm). .."
http://www.anandtech.com/show/7789/intel-talks-merrifield-moorefield-and-lte-at-mwc-2014
To prove my point, look no further than the CES that just happened. Intel chips are in approximately 3/4 of the newly announced tablets (actual tablets, not reference designs that will never be sold). I saw less ARM influence than ever before this year.
Intel paid about $1B to make that happen. They are putting extremely high incentives for the established PC system designers to use Intel parts instead of any of the ARM derivatives. That is in part a stop-gap measure. But is it also a bit necessary for folks targeting Android since it has historically had ARM designs , tools , and native code applied to Android systems. There is "extra" work if doing Android on Intel. Intel is buying into market share at this point.
When Android starts making major transition to 64-bit is going to be in much better position for the "buy in" that they are doing now. In short term though it also has side effect of killing off the Windows ARM systems.
Remember how much of a flop the Windows Surface RT was? Because nobody wanted to rewrite their apps for Windows on ARM?
It is not so much rewrite for Windows on ARM but to write to a brand new (non backwards compatible) Windows API. The new apps would largely run on either Windows 8 on x86 or ARM. The new API was a failure in part also because Windows Phone went no where fast. Much of the upshot was suppose to be could move apps across windows. The major downside was that only Windows OS with high moment/inertia was Windows on x86 and even that was heavily weighted to legacy x86 code inertia.
This new Windows API is in the slow adoption rate for the same reason it took MS until Windows 3.0 to get going full steam. Back in that transition DOS was the boat anchor inertia.
Rule of thumb is that it takes MS about 3 stabs at something to get it right.
If compare an original Surface Pro to the new Surface Pro 3 .... I doubt few would choose the former over the latter.
For the same reason, nobody will want to rewrite their apps for OS X on ARM.
OS X on ARM would not mean a new API. It does mean that legacy (currently owned software) isn't going to run without some solution to help. For PPC -> Intel Apple used a 3rd party company's solution. Transitive doesn't exist anymore and it has been almost two decades since Apple has done an architecture emulator..... I doubt those folks are still around at Apple since the talent wasn't around 10 years ago either.
The other huge problem is that Apple already
has an OS running on ARM. If Windows Phone was huge and growing we may not have seen Windows RT tablets that tried to more closely mimic Windows. MS was trying to use the leverage of Windows ( 90+% of classic PC market) to jumpstart an OS on ARM. [ one reason why Windows CE and others have had limited success in the past. ] Windows 10 is suppose to be a better context sensitive (which kind of device am I on) approach to convergence but that too is probably going to take a couple iterations to iron out. Remains to be seen if Windows 10 doesn't also mean that Windows ARM tablets mean the "phone" skew to Windows 10 and not the Surface (and similar x86 ones) variant.
Apple needs another OS on ARM like a another hole in the head. The OS with the bigger percentage of the "personal computer" market is iOS.
That is already on ARM. Apple sells about as many iPads in year as the whole Mac line up. The incremental addition of Minis would make no difference in the volume. It would only serve to shrink the discounts they do get from Intel on the CPUs that Apple's ARM implementation can not replace for the next couple of years. [ if ever, if Intel and AMD continues to make progress . ]
I think we'd sooner see an Apple tablet running x86 than we will see an Apple desktop running ARM.
General desktop or any desktop? ARM AppleTV was already done years ago. A small box AppleTV along same lines as Amazon FireTV and Google Nexus player has decent chance. Apple's Home Network API leverages AppleTV. A time capsule (back up the iOs devices at home) device that serves as a media server would not be surprising at all.
A keyboard less OS X laptop? I wouldn't hold my breath. A 360 hinge (fold keyboard away mode) with touchscreen would not be too surprising if sales of the Mac laptops go flat. Flat sales doesn't look like it going to happen any time soon. If the iPad Pro comes out and is successful even less likely.
A keyboard less iOS tablet on x86? Exceedingly doubtful. Apple does have enough resources already allocated to do a custom ARM chip specifically for higher end iOS. The volume is high enough ( several 10's of millions per year) that it is easy to keep that up for a long time.
Intel may start getting some radio orders again from Apple ( now that they are getting competitive in LTE solutions with world converage). But iOS application processors.... not really.
, only to be repurposed for the coming 2014 (and now 2015) minis.
