MacNN Forums (http://forums.macnn.com/)
-   MacNN Lounge (http://forums.macnn.com/macnn-lounge/)
-   -   A7 chip: is Apple the locomotive on this train? (http://forums.macnn.com/89/macnn-lounge/499624/a7-chip-is-apple-locomotive-train/)

 
fritzair Apr 10, 2013 03:45 PM
A7 chip: is Apple the locomotive on this train?
I remember when Apple was using the PPC back in the day. Jobs was dissatisfied when IBM, Motorola slowed development on the chips, the cycles became too long. Intel's cycles were much quicker and the chips from them were getting smaller. I think he became most upset when IBM developed the PPC chip for the XBox and the PS yet Apple was left out in the cold. His answer back then was to change to Intel and the rest is history.
Are we back there again? Will the A7 or someday the "A10" end up in the Mac desktop/portables? This chip can't be too far from being suitable for desktop and large mobile use.
If Apple does go this route I hope that they don't break the PC software capability built into our current desktops/portable with Bootcamp. I imagine VM or Citrix will fill in if needed.
Anyone else sensing this movement?
 
BLAZE_MkIV Apr 10, 2013 03:50 PM
The ARM chips are quite a ways from supplanting Intel chip in the non-mobile space. I see it as Apple developed their ARM version because there was no suitable chip available and no-one willing to tailor to there specific needs in such a tightly integrated environment. Laptops and certainly desktops don't need that level of integration due to higher power and space budgets. So it will be very hard to beat Intels economy of scale and improvement rate.
 
P Apr 10, 2013 04:40 PM
Intel's chips are very good right now, and Intel is clearly listening to Apple in their designs (the GT3e graphics option). Neither the CPUs nor the GPUs on the mobile side are anywhere near what Apple puts in their laptops, and Intel has a solid roadmap for both CPU and GPU development - and even with all that power, even the smallest 11" MBA has a 5 hour battery life. There is no need right now to trade performance for battery life.

For a comparison of the raw GPU power between laptops and mobiles, see this article.
 
OreoCookie Apr 12, 2013 04:31 AM
ARM-based CPUs will move in the direction of hotter, but faster chips, although that development will take years until we have an ARM CPU which is competitive with Intel's offering when it comes to per-core performance. And as P has said, Intel is working very, very hard to reduce the power consumption of their cpu + chipset combos, so they will encroach on ARM territory from above.
 
fritzair Apr 18, 2013 12:44 PM
More ammunition!
Quote, Originally Posted by fritzair (Post 4225631)
I remember when Apple was using the PPC back in the day. Jobs was dissatisfied when IBM, Motorola slowed development on the chips, the cycles became too long. Intel's cycles were much quicker and the chips from them were getting smaller. I think he became most upset when IBM developed the PPC chip for the XBox and the PS yet Apple was left out in the cold. His answer back then was to change to Intel and the rest is history.
Are we back there again? Will the A7 or someday the "A10" end up in the Mac desktop/portables? This chip can't be too far from being suitable for desktop and large mobile use.
If Apple does go this route I hope that they don't break the PC software capability built into our current desktops/portable with Bootcamp. I imagine VM or Citrix will fill in if needed.
Anyone else sensing this movement?
MacNN | Intel CEO pours cold water on possible Apple chip deal

This story touches on the issues I am trying to bring forward, and I assume that the "Lounge" is safe enough venue for the moderator.
Smarter, more savvy folks can chew on the above for a bit. Something is going on, based only on my experience with watching Apple, "one more thing..." could be a parting of ways from Intel. The above article doesn't come close to my statement, just I see it as Apple giving us a hint.
 
Spheric Harlot Apr 18, 2013 02:30 PM
I doubt that Apple would switch architectures for the Mac anytime before Windows loses its dominance. As the market for traditional PCs recedes and is eclipsed by the tablet/post-PC market, the increasing specialization of the PC's remaining market niches will allow specialized solutions and different architectures for those solutions.

But as long as the traditional PC is still a (or even the de facto) general computing platform, I don't see Apple switching away from intel/intel-compatible chips.
 
P Apr 19, 2013 09:18 AM
Quote, Originally Posted by fritzair (Post 4226654)
MacNN | Intel CEO pours cold water on possible Apple chip deal

This story touches on the issues I am trying to bring forward, and I assume that the "Lounge" is safe enough venue for the moderator.
Smarter, more savvy folks can chew on the above for a bit. Something is going on, based only on my experience with watching Apple, "one more thing..." could be a parting of ways from Intel. The above article doesn't come close to my statement, just I see it as Apple giving us a hint.
That's not what's happening here. That article (and the CNET piece as well) doesn't include the background, but it's basically like this: Intel changes its production process every other year. Every time they change, they switch to a smaller node size, which among other things lets you put twice as many transistors in the same area. This has gone on for decades. What is recent is that Intel can no longer find a use for all those transistors - they've integrated a lot of what used to be other chips on the motherboard, and they've increased the cache as far as makes any real sense. Future chips are smaller than the older ones. Haswell, the next generation, brings in the entire southbridge into the package and the VRMs onto the die itself. After that, there's only RAM and the very analog chips left outside the CPU, and Intel is working on the analog stuff as well. At the same time, the PC market has started shrinking - slowly yet, but accelerating.

This means that Intel has a lot of spare capacity. At the same time, Intel's foundries are the best in the business, and Intel can charge customers through the nose for that capacity. Only makes sense to start making chips for high-demanding customers, right? Intel has begun doing that, but only in areas where it does not compete, like FPGA and I think some networking hardware.

The question now is what Intel will do in the future. Will they make use of their plants to launch an upscale foundry business - the short term option - or do they remain focused on breaking in to the tablet and phone business? In the first case, Apple would be the perfect customer - in the second, Intel shouldn't let any competitor use its foundries, and especially not a high-profile one like Apple. So far, Intel sticks with the long term plan, but that may not last if they don't start bringing home some design wins for their chips some time soon.

The reason people keep asking about Apple using Intel capacity isn't so much because they're interested in what Apple will do - it really doesn't have anything to do with Apple. They do it as a proxy question for what Intel's future plans are. If Intel were best buddies with Google and Googlorola were making the top-of-the-line gear, they'd ask the same question about Google and Intel.

There is nothing in that piece about Intel chips in the Macs. Personally I think that the moment to dump Intel in Macs is past - if they would have done it, it was back in the Arrandale era, when Intel's graphics were so bad that Apple preferred to stay with Core 2. Now with Haswell bringing significantly better graphics, Apple has received most of what it wanted. I think Apple is very happy with Intel's plans for Haswell and Broadwell (and given that they're Apple, they probably know a bit about Skylake and Skymont and probably a few codenames beyond that).
 
All times are GMT -4. The time now is 01:56 PM.

Copyright © 2005-2007 MacNN. All rights reserved.
Powered by vBulletin® Version 3.8.8
Copyright ©2000 - 2014, vBulletin Solutions, Inc.


Content Relevant URLs by vBSEO 3.3.2