Apple decision about Skylake versus Broadwell microprocessors delaying MBA Retina?

[Republius]

The Intel Skylake microprocessor likely will not be available until mid-2015. But the Intel Broadwell Y Core M microprocessor that is supposed to be available in mass quantities by the beginning of 2015 seemingly will have very little speed improvement over the previous generation Haswell microprocessors, though it will allow for super thin devices due to how energy efficient they are (thus eliminating the need for an internal fan) and be able to drive a retina screen at the same time.

If Apple manufactures a MacBook Air Retina with the Broadwell Y Core M microprocessor, it may not be much of a technological advancement other than in slimness and the addition of the retina screen. But if Apple waits to manufacture a MacBook Retina with the next generation Skylake microprocessor, there will arguably be speed enhancements to go along with a super thin profile and retina screen.

Is this why the MacBook Air Retina has been delayed? To a layman it sounds as if Apple ought to wait another half-a-year for the Intel Skylake microprocessor for this new device rather than succumb to using the Intel Broadwell Y Core M microprocessor. And I say this as someone who needs and is ready to purchase a MacBook Air Retina now.

 

[FrozenDarkness]

Macbook air hasnt really been the place for processor speed bumps. Graphics is whereits important

 

[Republius]

Macbook air hasnt really been the place for processor speed bumps. Graphics is whereits important

Understood and acknowledged.

But given that the delays in manufacturing Broadwell (which also arguably speak to its being a problematic product) have bumped it up against release of the following generation Skylake, why not wait a few more months, if you are Apple, for Skylake in order to be able to produce a truly breakthrough laptop that marks significant advancement in slimness, screen clarity, and speed?

As a matter of theory, a MacBook Air that merely adds retina and thinness with no parallel upgrade in speed even compared to the previous generation Haswell microprocessor may well be properly deemed a boring and unworthy iteration in the line.

And, as a matter of fact, if Apple was going to go with Broadwell Y Core M in the MacBook Air Retina then there is no reason for it to not have been announced by now given that they should have had first call on such microprocessors in the face of the Lenovo Yoga 3 Pro (with the Broadwell Y Core M) having been announced this month and readied for release in November. It should also be noted, as a factual matter, that Intel did not name Apple as among those companies building upcoming products with the Broadwell Y Core M microprocessor when it was announced.

So it would make sense if Apple waiting for Skylake was the reason for the delay in announcing and making available a MacBook Air Retina. And it seems, to me, to be a good reason for delay - as the product will be substantially better and certainly more cutting edge using Skylake a few months hence.

 

[orestes1984]

simple really, wait for a real speed bump and then upgrade, otherwise you're just wasting money.

 

[Obi Wan Kenobi]

I think your analysis is sound, Republius. Looks like we have a wait until the Skylake chips are ready.

 

[FrozenDarkness]

I don't know much about timelines, all I know is Apple will most likely refresh the macbook air before the summer as is their yearly tradition and people will be disappointed about the lack of a retina display. My own theory is that we'll see a 12" and a 14" rmba and a higher resolution rmbp 15" with a 4K display later that year.

But on a practical matter, Haswell processors aren't much faster than the Ivy Bridge ones. Apple still released it becuase Haswell brought the "all day" battery life. So on history, I disagree with you on needing a spec bump. If apple were to release a rmba, we just need a graphics bump with enough vram to handle the higher res screen.

 

[Gildarts]

I've been thinking something similar. I wanted to register on MacRumors just to pitch this theory. I let the idea slide but it seems I'm not the only one thinking this. I'm glad you pointed this out.

The MBA is usually upgraded at WWDC. If they follow that path next year it's unlikely that the MBA will have an Intel Skylake processor. But releasing a retina MBA Broadwell at that time would be suicide. It would be DOA (at least to me).

I'm sure that in the fall all the MBP's will be getting the Intel Skylake processors.

Considering Intel's tock-tick release strategy with Broadwell being the tick and Skylake being the tock it would make sense to just skip Broadwell altogether.

Skylake is going to be bringing Thunderbolt 3, WiGig, Wireless charging, fanless design, 4K graphics, and tons of other stuff.

Broadwell just brings better graphics and a fanless design.

I'm still debating on MBA vs. MBP but I'm definitely, undoubtedly, skipping Broadwell.

 

[Gildarts]

I just checked WWDC 2014. There were no hardware announcements. Completely software focused.

The initial release of Haswell MBA's was in june 2013. There was a small refresh (updated processors, also Haswell) in april but nothing huge. If Apple continues it's current trend then there will be nothing announced at WWDC 2015. It would just be software focused (standard iOS & OS X).

BUT, there's still the watch. I don't know if they're planning to give another keynote Q1 2015. I'm REALLY hoping that if they do they don't announce anything Broadwell.

Lastly, Intel has specifically stated that Skylake is on schedule. So why waste any time on Broadwell? That ship has sailed.. Skylake is going to be epic. Which means Apple can be epic. Broadwell is great, but not epic (except for maybe the fanless design, which is also coming to Skylake with tons of other crazy innovations).

 

[Crazy Badger]

I'm hoping for a MBA with retina screen before October 2016 when I'm due to replace my 2013 model

 

[motrek]

The Intel Skylake microprocessor likely will not be available until mid-2015. But the Intel Broadwell Y Core M microprocessor that is supposed to be available in mass quantities by the beginning of 2015 seemingly will have very little speed improvement over the previous generation Haswell microprocessors, though it will allow for super thin devices due to how energy efficient they are (thus eliminating the need for an internal fan) and be able to drive a retina screen at the same time.
...

Intel has done a lot of talking about thin fanless designs but it didn't seem to pan out for this generation:

http://ultrabooknews.com/2014/10/16...erformance-tests-reveal-throttling-fan-noise/

I wouldn't expect a fanless MacBook Air in the next few years (or ever?) or one that's significantly thinner.

 

[Gildarts]

From what I understand the fanless design depends on which chipset. The Broadwell core M being like the current core i3 would not allow for a fanless design, but higher-end chipsets (within the Broadwell product) would. Intel seems to be rather confusing about this. They have boasted the "fanless" capabilities of Broadwell whenever they could. If it can't.

All the more reason to skip Broadwell. Although Skylake will still be 14nm it brings some big innovations in terms of I/O, power and graphics.

 

[motrek]

From what I understand the fanless design depends on which chipset. The Broadwell core M being like the current core i3 would not allow for a fanless design, but higher-end chipsets (within the Broadwell product) would. Intel seems to be rather confusing about this. They have boasted the "fanless" capabilities of Broadwell whenever they could. If it can't.

All the more reason to skip Broadwell. Although Skylake will still be 14nm it brings some big innovations in terms of I/O, power and graphics.

I've posted this before but it was a couple months ago. Basically, a 30% power reduction vs. Haswell (which is what Intel is claiming for Broadwell) is not nearly enough to produce a fanless, high-performance laptop. The Yoga proves this point.

But that's no reason to be down on Broadwell. I'm sure it's a fine chip that will increase battery life and (graphics?) performance vs. Haswell.

 

[Gildarts]

I'm sure it's an improvement on Haswell but from what I'm getting it's not worth the upgrade. Let's be real, tick's are best skipped for tocks.

I'm reading the link that you sent me and it seems to confirm my suspicions. See the excerpts below.

I’ve been in touch with Mobilegeeks.de who are testing the Yoga 3 Pro now and have written up their first impressions (in German) I wanted to find out more about the fan. Yes, a fan is included with the Yoga 3 Pro. In some respects that will be a good thing as it increases the thermal ‘space’ for Turbo Boost to work. Over 50% of the CPU performance of Core M relies on their being enough scope for heating up as it overclocks. If the temperature is already too high or rises too quickly, Tube Boost can turn off and you’re left with a 1.1Ghz CPU which has nothing like the power of the previous Yoga 2 Pro.

Roland, the reviewer over at Mobilegeeks, points me to a forum thread on Notebookreview where there are benchmarks and evidence of throttling in tablet mode and in multi-threaded CPU tests. This confirms my worry about Core M. Intel can show nice high-speed tests in optimized casings but it’s up to the manufacturer to create the balance between size and performance. Having said that it’s disappointing that even with a fan and the high-end Core M 5Y70 the Yoga 3 Pro is not performing like and Ultrabook. In the Mobilegeek Cinebench 11.5 multi-cpu test the Yoga 3 Pro scored just 2.08 which is less than the Lenovo Yoga 11S with the Core i7 Y-series from last year and less than the original Lenovo Yoga 13. This is not a good test result. Here’s the performance figure slotted into our Ultrabook performance table. Note that the cheaper Surface Pro 3 wih Core i5 U-series is going to bring you nearly 50% more CPU power.

The Mobilegeeks first impressions review is here. (translated) and you’ll find positive comments about the keyboard and build, the screen and the weight. There are also positive comments about battery life which can go up to 9 hours in video playback mode but there are definitely issues to consider here.

 

[Republius]

My thanks to all of those who have pointed out that using the Intel Broadwell Y Core M microprocessor does not necessarily lead to a design of a potential MacBook Air Retina that can eliminate an internal fan. If so, all the more reason for Apple to wait until Skylake is available before mass producing the next generation MacBook Air with retina display.

I should have also mentioned that most of the informed leaks about the MacBook Air Retina (from John Paczkowski, Jack March, Ming-Chi Kuo, etc.) do not mention what kind of microprocessor this new product iteration will utilize, which seems significant in terms of a key component not being leaked.

I am a layman who is trying to learn. In attempting the switch from Windows to Mac, I am convinced that the MacBook Air is my best bet. And I am incredibly anxious for an upgraded model with retina. But everything I read leads me to believe that a MacBook Air Retina using Broadwell Y Core M as the microprocessor, which could be made available now, would be disappointing and immediately behind the technology curve. Apple might as well wait a few months longer and put Skylake microprocessors in the MBA Retina.

The next question is what such would do to the MacBook Pro line? Perhaps the MacBook Air line becomes the super mobile choice with its very thin design with the MacBook Pro becoming the desktop replacement choice with more power by retaining some bulk?

 

[foobarbaz]

I don't get the excitement about Skylake. Intel's "tick" (i.e. Broadwell) is where power savings happen. That's when they'll shrink the rMBP to Air size.

Skylake will certainly add power to the machines, but it won't be a revolutionary step by a long shot. There isn't a machine that Apple can build with Skylake that it can't build with Broadwell.

Even if Skylake is released just 3 months (and I think it will be 6 at the very least) after Broadwell, that's still a full business quarter in which to sell machines. Waiting even longer gains them nothing.

 

[Gildarts]

I don't get the excitement about Skylake. Intel's "tick" (i.e. Broadwell) is where power savings happen. That's when they'll shrink the rMBP to Air size.

Skylake will certainly add power to the machines, but it won't be a revolutionary step by a long shot. There isn't a machine that Apple can build with Skylake that it can't build with Broadwell.

Even if Skylake is released just 3 months (and I think it will be 6 at the very least) after Broadwell, that's still a full business quarter in which to sell machines. Waiting even longer gains them nothing.

From what I understand from Intel. Skylake is when Intel and "the industry" will do away with wires. WiGig, Wireless charging, Wireless connecting to displays (without adapter).

So while Broadwell will bring graphics performance and battery savings. Skylake seems to be doing that and more.

 

[Republius]

I don't get the excitement about Skylake. Intel's "tick" (i.e. Broadwell) is where power savings happen. That's when they'll shrink the rMBP to Air size.

Skylake will certainly add power to the machines, but it won't be a revolutionary step by a long shot. There isn't a machine that Apple can build with Skylake that it can't build with Broadwell.

Even if Skylake is released just 3 months (and I think it will be 6 at the very least) after Broadwell, that's still a full business quarter in which to sell machines. Waiting even longer gains them nothing.

I get that nothing is certain in this business until and unless the benchmarks come out. But absent having them in this instance there may be a few other things to consider. One is that Broadwell has been problematic from the get-go and is going to end up being the Intel microprocessor with the shortest lifespan ever. The Lenovo Yoga 3 Pro is already manifesting issues because of it. Why would Apple want to build a breakthrough product on its architecture?

In addition, Intel representatives are claiming Skylake will provide the biggest personal computer advancements in the last 10 years, and will reflect a "significant increase in performance, battery life, and power efficiency." That sounds like a microprocessor Apple would want to build its next generation MacBook Air around.

Please see the following article for a thorough treatment of Broadwell versus Skylake: http://www.pcworld.com/article/2683392/pc-confusion-to-linger-on-intels-quick-jump-to-skylake.html

The bottom line is that if Apple was going to come out with a Broadwell laptop it would have already. But because Broadwell is developing as a quirk, problematic iteration for Intel with something significantly better around the corner, it seems prudent for Apple to wait for Skylake.

 

[Woochoo]

But there's something I don't understand (if someone can make it clear it to me). Intel delayed Broadwell for a long, but despite Skylake not being in "the same production pipeline" (so it wouldn't be directly delayed due to Broadwell problems), why wouldn't Intel delay Skylake? They will need to get rid of all the Broadwell stock, I suppose, no?
I mean, will they release Skylake "at time" as it was planned?

 

[Republius]

But there's something I don't understand (if someone can make it clear it to me). Intel delayed Broadwell for a long, but despite Skylake not being in "the same production pipeline" (so it wouldn't be directly delayed due to Broadwell problems), why wouldn't Intel delay Skylake? They will need to get rid of all the Broadwell stock, I suppose, no?
I mean, will they release Skylake "at time" as it was planned?

Please see this article for insights as to your concerns: http://www.pcworld.com/article/2683392/pc-confusion-to-linger-on-intels-quick-jump-to-skylake.html

 

[leonw]

There are more Broadwell cpu's on the way. The new mba is not going to have a core M. It looks like it's not powerful enough (lenovo yoga 3) and would have been anounced at the last Apple event if it was.
But why could the next mba not have a more powerful Broadwell to run a retina screen. It might use more power than the core M but still less than Haswell.
And there might be savings on improved battery technology and what happened to those igzo screens that use less power?
Apple never uses low end cpu's, I've heard. They might well introduce a new mba with Broadwell ( apple watch event early 2015?) and an upgrade with Skylake later that year. It gives them time to iron out the problems of that first gen rMBA. If they wait for the next better generation cpu, they will never release anything. There is always a better next on its way.

 

[Republius]

There are more Broadwell cpu's on the way. The new mba is not going to have a core M. It looks like it's not powerful enough (lenovo yoga 3) and would have been anounced at the last Apple event if it was.
But why could the next mba not have a more powerful Broadwell to run a retina screen. It might use more power than the core M but still less than Haswell.
And there might be savings on improved battery technology and what happened to those igzo screens that use less power?
Apple never uses low end cpu's, I've heard. They might well introduce a new mba with Broadwell ( apple watch event early 2015?) and an upgrade with Skylake later that year. It gives them time to iron out the problems of that first gen rMBA. If they wait for the next better generation cpu, they will never release anything. There is always a better next on its way.

The Broadwell U microprocessor comes out at CES in January. But it is still only modestly better than its Haswell predecessor and part of the problematic Broadwell iteration. Why not wait a few more months, if you are Apple, to utilize Skylake in the MacBook Air Retina given that it is far superior to Broadwell, has experienced a smooth and on-time gestation, and will provide cutting-edge performance?

Again, because Broadwell is bumping up against Skylake, the former will be passe very shortly after it can be utilized in the mass market. What Apple is seemingly trying to accomplish with the MacBook Air Retina would be much better served utilizing Skylake after just a few extra months of waiting.

 

[motrek]

The Broadwell U microprocessor comes out at CES in January. But it is still only modestly better than its Haswell predecessor and part of the problematic Broadwell iteration. Why not wait a few more months, if you are Apple, to utilize Skylake in the MacBook Air Retina given that it is far superior to Broadwell, has experienced a smooth and on-time gestation, and will provide cutting-edge performance?

Again, because Broadwell is bumping up against Skylake, the former will be passe very shortly after it can be utilized in the mass market. What Apple is seemingly trying to accomplish with the MacBook Air Retina would be much better served utilizing Skylake after just a few extra months of waiting.

Remember, Apple released the 2010 MBAs with Core 2 Duos when they were already basically obsolete and then switched to i5s and i7s about ~6 months later.

The big news about a redesigned MBA will be its form factor and its screen (presumably retina), not which processor it's using.

 

[Republius]

Remember, Apple released the 2010 MBAs with Core 2 Duos when they were already basically obsolete and then switched to i5s and i7s about ~6 months later.

The big news about a redesigned MBA will be its form factor and its screen (presumably retina), not which processor it's using.

Of course in this particular instance the microprocessor will have significant impact on the form. Ostensibly Apple is trying to arrive at a super slim profile for the MacBook Air Retina, which can only be manifest with a microprocessor that eliminates the need for an internal fan. In addition, I would imagine Apple regrets, in hindsight, outfitting the 2010 MacBook Air with obsolete microprocessors - or at least I hope doing such is something a reconstituted organization under Tim Cook will choose to avoid in pursuing product value.

As an aside, given the weight Apple carries in the personal computer world, which is second-to-none, why aren't they in the habit of being the first manufacturer in the market with a product utilizing a brand new microprocessor?

I recognize that there is some risk in doing such, as Lenovo is experiencing with the Yoga 3 Pro and the Intel Broadwell Y Core M microprocessor. But from all reports the Skylake microprocessor has had a very smooth and on-time gestation period - and it seems tailor-made for a slimmer and faster MBA Retina.

As badly as I want to be among the first to obtain a MacBook Air Retina, what I am finding would cause me to hesitate if they build it with a Broadwell microprocessor - which may turn out to be a lemon for Intel. I don't think Broadwell is as efficient in terms of heat and speed as it needs to be, which is why I am wondering if Apple may eschew it and wait for Skylake for the MBA Retina.

 

[rekhyt]

I've been thinking something similar. I wanted to register on MacRumors just to pitch this theory. I let the idea slide but it seems I'm not the only one thinking this. I'm glad you pointed this out.

The MBA is usually upgraded at WWDC. If they follow that path next year it's unlikely that the MBA will have an Intel Skylake processor. But releasing a retina MBA Broadwell at that time would be suicide. It would be DOA (at least to me).

I'm sure that in the fall all the MBP's will be getting the Intel Skylake processors.

Considering Intel's tock-tick release strategy with Broadwell being the tick and Skylake being the tock it would make sense to just skip Broadwell altogether.

Skylake is going to be bringing Thunderbolt 3, WiGig, Wireless charging, fanless design, 4K graphics, and tons of other stuff.

Broadwell just brings better graphics and a fanless design.

I'm still debating on MBA vs. MBP but I'm definitely, undoubtedly, skipping Broadwell.

Oh wow.

I definitely wasn't aware of the massive round of changes with Skylake.

In terms of 4K graphics... ... is this to do with the integrated graphics (support for 4K?) or something else? I was under the impression that graphics card capability was the only thing that prevents 4K.

 

[FrozenDarkness]

Oh wow.

I definitely wasn't aware of the massive round of changes with Skylake.

In terms of 4K graphics... ... is this to do with the integrated graphics (support for 4K?) or something else? I was under the impression that graphics card capability was the only thing that prevents 4K.

I mean is there really an indication on how that's how Intel is releasing Skylake? Some rumors point to thunderbolt 3 not coming out until 2016.

I've read that Intel's tick tock strategy is also different for graphics, with a bump in graphics in the tick part and the processing in the tock part. If broadwell is enough to drive a retina 1440x900 display (2880x1800) than it'd be a huge upgrade and apple doesn't need to wait for skylake at all.