Apple decision about Skylake versus Broadwell microprocessors delaying MBA Retina?

[heliora]

my experience with my rMBP 13-inch suggests that more power is needed to drive the Retina display smoothly, any moderate to heavy usage causes the system to lag.

i'm only a layperson in all these so i could be wrong, MBA is all about lightness, portability and battery life, and that means making trade-offs with a lower-res screen, because a Retina display necessitates higher specs and power which results in a bigger profile, weight and shorter battery life.

basically currently if you want MBA with a Retina display you basically end up with a rMBP

so i think the primary concern for Apple in developing the MBA is maintaining the portability and battery life, and the reason they haven't introduced Retina display is because they haven't found a way to do that without sacrificing either portability or battery life.

if Broadwell allows them to do that, then they will do it, but my guess is it doesn't, i don't think it's powerful or energy-efficient enough to drive a Retina display without any sacrifice in portability or battery life in MBA

and if so it begs the next question, if they update the MBA with Broadwell without introducing Retina display, that means the MBA would have even better battery life, which brings the same problem for their subsequent update, i don't think Apple ever regresses on a certain battery life achieved, at worst they would wanna maintain it, so if there's a better battery life with Broadwell, an update with Retina display must at least maintain that, would Skylake achieve that?

or Apple can skip Broadwell entirely and go straight to Skylake which is supposed to be more powerful to drive a Retina display while maintaining the current battery life in MBA

so if what i said about Broadwell is true, then my conjecture is there'll not be any update to MBA until Skylake is available, and it will have a Retina display.

which then makes me wonder what will MBP be like, but that's another topic.

 

[Republius]

It is becoming more evident that if Apple uses Broadwell in a MacBook Air Retina it will be the U rather than Y Core M chip given that the latter does not seem to be able to provide fanless design (as can be gleaned from the Lenovo Yoga 3 Pro) that will supposedly be part of the next MBA iteration. The Broadwell U chip is coming out at CES in January reportedly and will be available for mass distribution in February.

Of course if Skylake can be shipped by second quarter 2015, as some are saying is possible, then Apple will have to make a choice - though such could be moot if Skylake will do things essential to the Apple vision for the MBA Retina that Broadwell U cannot.

I am not technologically proficient enough to discern between the two, but obviously Skylake represents the next generation and is claimed by Intel to be a significant (a tock in their nomenclature compared to the Broadwell tick) advancement in microprocessors.

So my guess would be that if Broadwell U can effectuate the Apple vision for a MacBook Air Retina - which from most reports seems to be fanless design, retina display, no reduction in processing power, and minimal battery life loss - then they will use it to market a device in early 2015 that they had been hoping to make available by late 2014 and use Skylake in the next iteration of MacBook Pros for Fall of 2015, which would keep the product lines differentiated.

 

[motrek]

...
So my guess would be that if Broadwell U can effectuate the Apple vision for a MacBook Air Retina - which from most reports seems to be fanless design, retina display, no reduction in processing power, and minimal battery life loss - then they will use it to market a device in early 2015 that they had been hoping to make available by late 2014 and use Skylake in the next iteration of MacBook Pros for Fall of 2015, which would keep the product lines differentiated.

The only people who are expecting a fanless design are the people who have bought into Intel's hype too much and who generate unreliable Apple rumors.

There has been no credible indication that Apple wants to make a fanless laptop.

Personally I don't even understand why some people might want this. Is the tiny, inaudible fan in the current MBAs really so upsetting?

 

[leonw]

The only people who are expecting a fanless design are the people who have bought into Intel's hype too much and who generate unreliable Apple rumors.

There has been no credible indication that Apple wants to make a fanless laptop.

Personally I don't even understand why some people might want this. Is the tiny, inaudible fan in the current MBAs really so upsetting?

Agreed.
There will be fanless designs in the (near?) future but it doesn't seem to be there yet. Maybe not even Skylake is powerful enough. I don't know.
If so, Broadwell and Skylake could have the same form factor and there could be a Broadwell rMBA first which will be upgraded to Skylake late 2015 with better performance, battery life and issues of the first gen. solved.

 

[Gildarts]

Meh. I'm not super impressed with the MBP's specs vs. the MBA. The MBA has enough thermal headroom to run its CPU at maximum clock speed indefinitely. So you're basically talking about a 3.1GHz CPU vs. a 2.7GHz CPU if we're considering base models. That 13-14% difference isn't enough to get me too excited.

Now if you're doing something that requires a bunch of RAM or a bunch of GPU power then that's a different story, but most things don't.

So I'm not surprised at all if some people are claiming that the MBA performs similarly to the MBP, because it does, regardless of retina display or whatever.

The MBA does not have enough thermal headroom to run the CPU at maximum clock speed indefinitely. I would know, I have one.

As far as I know, pushing a machine to the limit means making it use all of its resources (which includes RAM and GPU) at full capacity.

So my point still stands.

 

[Gildarts]

so i think the primary concern for Apple in developing the MBA is maintaining the portability and battery life, and the reason they haven't introduced Retina display is because they haven't found a way to do that without sacrificing either portability or battery life.

if Broadwell allows them to do that, then they will do it, but my guess is it doesn't, i don't think it's powerful or energy-efficient enough to drive a Retina display without any sacrifice in portability or battery life in MBA

and if so it begs the next question, if they update the MBA with Broadwell without introducing Retina display, that means the MBA would have even better battery life, which brings the same problem for their subsequent update, i don't think Apple ever regresses on a certain battery life achieved, at worst they would wanna maintain it, so if there's a better battery life with Broadwell, an update with Retina display must at least maintain that, would Skylake achieve that?

or Apple can skip Broadwell entirely and go straight to Skylake which is supposed to be more powerful to drive a Retina display while maintaining the current battery life in MBA

so if what i said about Broadwell is true, then my conjecture is there'll not be any update to MBA until Skylake is available, and it will have a Retina display.

which then makes me wonder what will MBP be like, but that's another topic.

You make an obvious yet compelling and logical case. I don't know why I didn't approach the subject from this point of view earlier. It's so obvious.

I'm now even more convinced that they're skipping Broadwell.

 

[bjet767]

Republios

You make some interesting points about speed, delays and such, but you forget the most important part; Apple is not in the game to win speed tests.

Apple has never dominated the computer speed world nor it appears ever attempted to. When I review the specs of all the hardware Apple produces one can always easily find more bang for the buck (and sometimes substantially so) in some form of Windows or Android based OS system.

We/I buy Apple for the quality, reliability and simplicity of their design and function. Of course many just do it for the "cool" and cultist factor. All, however, of us pay a premium for these features and most are happy to do so and that is what Apple banks on.

You also forget the newest iPads, iPhones, iMacs, and the coming Airs weren't developed overnight. This newest Air has probably been in the works for at minimum three years and to just delay it for the next processor is fool hardy at best. This is a Corporation which needs to make money and development costs must have a financial return.

Even if there isn't any improvement in speed I will upgrade from my 2014 11" Air to the newest one if it offers a retina screen, more visual real estate, longer battery life and especially my dream Mac feature, a touch screen.

----------

BTW many seemed to forget the rule of computing:

There will always be a better, faster, leaner, more powerful, etc, processor just six month away.

No Apple won't wait for the next best processor. It is probably already designing and planning the next generation Air for 2017.

Don't you think Intel works very close with Apple as it develops new processors? They don't design new CPUs just because they can.

 

[motrek]

The MBA does not have enough thermal headroom to run the CPU at maximum clock speed indefinitely. I would know, I have one.

Are you using your Mac in Death Valley? Because at room temperature, I've had my 11" MBA under full load for hours and the fastest the fan ever spun was 4000 RPM and it never wavered from full turbo boost frequency.

If your MBA can't sustain full turbo boost then you should be looking for the reason why not. (Or maybe you don't have a Haswell MBA?)

As far as I know, pushing a machine to the limit means making it use all of its resources (which includes RAM and GPU) at full capacity.

So my point still stands.

Different people do different things with computers and it's possible to "push a computer to the limit" in different ways... you could run cutting edge 3-D games and hit the graphics limit before using a fraction of your CPU power or RAM... or you could do movie editing and hit your RAM limit way before any CPU or graphics limit... or you could do something compute-intensive and not involve graphics or RAM. And disk performance could be thrown into this equation for good measure.

I will agree that if you have no idea what the computer will be used for, a MBP is a smart choice since it will be between "slightly faster" and "much faster" at everything.

But for the things that most people typically do with computes, and a lot of things that are atypical, the MBA will effectively have the same performance as a MBP just because its CPU runs more or less as fast.

 

[Gildarts]

Are you using your Mac in Death Valley? Because at room temperature, I've had my 11" MBA under full load for hours and the fastest the fan ever spun was 4000 RPM and it never wavered from full turbo boost frequency.

If your MBA can't sustain full turbo boost then you should be looking for the reason why not. (Or maybe you don't have a Haswell MBA?)

But for the things that most people typically do with computes, and a lot of things that are atypical, the MBA will effectively have the same performance as a MBP just because its CPU runs more or less as fast.

I don't have a Haswell MBA. I have an 11" 2012 MBA (Ivy Bridge).

Different people do different things with computers and it's possible to "push a computer to the limit" in different ways... you could run cutting edge 3-D games and hit the graphics limit before using a fraction of your CPU power or RAM... or you could do movie editing and hit your RAM limit way before any CPU or graphics limit... or you could do something compute-intensive and not involve graphics or RAM. And disk performance could be thrown into this equation for good measure.

I never disputed this. I merely gave my definition of "pushing to the limit" and argued based on that.

But for the things that most people typically do with computes, and a lot of things that are atypical, the MBA will effectively have the same performance as a MBP just because its CPU runs more or less as fast.

Again, this was not the point being made by Republius and never the point of argument.

The point was that the difference in specs was solely because of the retina display in the MBP and that because of that (the retina display) the performance is equal between the two products (MBA vs. rMBP).

 

[motrek]

I don't have a Haswell MBA. I have an 11" 2012 MBA (Ivy Bridge).

Okay. I didn't mean that every MBA in history had enough thermal headroom to prevent throttling, just the current (and presumably future) ones.

Again, this was not the point being made by Republius and never the point of argument.

The point was that the difference in specs was solely because of the retina display in the MBP and that because of that (the retina display) the performance is equal between the two products (MBA vs. rMBP).

We will have to agree to agree on this very literal point then.

But I will continue to maintain my related point which is that it should be no surprise if the MBA performs as well as the MBP at many/most things since they have very similar hardware.

 

[Republius]

The only people who are expecting a fanless design are the people who have bought into Intel's hype too much and who generate unreliable Apple rumors.

There has been no credible indication that Apple wants to make a fanless laptop.

Personally I don't even understand why some people might want this. Is the tiny, inaudible fan in the current MBAs really so upsetting?

Are you even researching what is being forecast for the next MacBook Air? What you posit could not be further from the truth.

Once again, for all of those who question what will be in the next iteration of MacBook Air, reliable sources indicate that a 12-inch retina screen, fanless design (leading to an even slimmer profile), and Intel microprocessor are givens. Of course we will not know for sure until the product is announced. But the quantity and quality of the sources seemingly make discussions absent assuming those three features sophistry. Those claiming inside information about the next MacBook Air are all saying the same thing: it will have a 12-inch retina screen and be super slim due to a fanless design and run on an Intel chip.

Jack G. March has been completely correct about recent Apple product releases based on his sources. And those sources indicate a 12-inch retina display and fanless design with an Intel microprocessor are coming (http://jackgmarch.com/2014/09/22/exclusive-12-macbook-air-design-details/) in the next MacBook Air. And multiple additional sources are supporting his report (http://9to5mac.com/2014/03/21/rumor...on-without-fan-or-mechanical-trackpad-button/).

How Apple will get all of this done together may be questionable given that no Intel microprocessor yet tested seems capable of it. But these key design features of a new MacBook Air are simply not in dispute among those who claim inside sources as to the device and have a track record of reliability in reporting on upcoming Apple product releases.

 

[motrek]

Are you even researching what is being forecast for the next MacBook Air? What you posit could not be further from the truth.
...

Heh, yes, I have seen all the rumors. Whoops, sorry, I mean I have "researched" the "forecasts."

The difference is that I've been around the block and I've seen a million rumors from so-called "reputable sources" with "inside contacts" not pan out. So until Apple says something themselves or leaks something to the WSJ, you'll excuse me if I don't buy in.

 

[Gildarts]

Okay. I didn't mean that every MBA in history had enough thermal headroom to prevent throttling, just the current (and presumably future) ones.



We will have to agree to agree on this very literal point then.

But I will continue to maintain my related point which is that it should be no surprise if the MBA performs as well as the MBP at many/most things since they have very similar hardware.

I acknowledge your point and I do not dispute/refute it.

----------

Are you even researching what is being forecast for the next MacBook Air? What you posit could not be further from the truth.

Once again, for all of those who question what will be in the next iteration of MacBook Air, reliable sources indicate that a 12-inch retina screen, fanless design (leading to an even slimmer profile), and Intel microprocessor are givens. Of course we will not know for sure until the product is announced. But the quantity and quality of the sources seemingly make discussions absent assuming those three features sophistry. Those claiming inside information about the next MacBook Air are all saying the same thing: it will have a 12-inch retina screen and be super slim due to a fanless design and run on an Intel chip.

Jack G. March has been completely correct about recent Apple product releases based on his sources. And those sources indicate a 12-inch retina display and fanless design with an Intel microprocessor are coming (http://jackgmarch.com/2014/09/22/exclusive-12-macbook-air-design-details/) in the next MacBook Air. And multiple additional sources are supporting his report (http://9to5mac.com/2014/03/21/rumor...on-without-fan-or-mechanical-trackpad-button/).

How Apple will get all of this done together may be questionable given that no Intel microprocessor yet tested seems capable of it. But these key design features of a new MacBook Air are simply not in dispute among those who claim inside sources as to the device and have a track record of reliability in reporting on upcoming Apple product releases.

Well this is interesting (9to5): a completely redesigned trackpad and it's coming in Space Gray, Gold and Silver.

Forgive me but I'll disregard everything in the 9to5 article since the source was completely wrong about date and release times.

This is also very interesting:

The source also believes that the MagSafe port as we know it has also been sacrificed and the new Macbook Air charges in a different way to what we’ve seen in current macs, however the source wasn’t too sure whether this had been finalised or not.

Not to mention this:
As I reported last week, the product is planned to arrive in Space Grey, Silver, and Gold to launch in Mid 2015.

Space Gray might be very nice..

 

[Gildarts]

Upon reading the comments at 9to5mac at the link Republius provided it seems that this 12 inch macbook might not be a new Air but a whole new line of MacBooks.

This nuance seemed to have gotten lost in translation from chinese to english.

Very interesting. Although to me it does not make sense. I mean, Apple being all post-pc era and stuff. Why introduce a new line of MacBooks. Not that I'm complaining..

Just interested in the reasoning.

 

[Republius]

I acknowledge your point and I do not dispute/refute it.

----------



Well this is interesting (9to5): a completely redesigned trackpad and it's coming in Space Gray, Gold and Silver.

Forgive me but I'll disregard everything in the 9to5 article since the source was completely wrong about date and release times.

This is also very interesting:

The source also believes that the MagSafe port as we know it has also been sacrificed and the new Macbook Air charges in a different way to what we’ve seen in current macs, however the source wasn’t too sure whether this had been finalised or not.

Not to mention this:
As I reported last week, the product is planned to arrive in Space Grey, Silver, and Gold to launch in Mid 2015.

Space Gray might be very nice..

I would not assume these sources got the release dates of the MBA Retina wrong. There seems to be plenty of evidence that Apple wanted to release the device for 2014 holiday sales. But what everybody (Apple, Intel, analysts, and sources) got wrong was the availability of the Intel chip to allow for fanless design that can drive retina without compromising processing power and battery life.

As a parenthetical matter, I will get the MacBook Air Retina regardless. But it is a bit squeamish to consider that the device will be using a relatively untested microprocessor (Broadwell U or Skylake). I hope Apple and Intel get this right.

 

[Gildarts]

I would not assume these sources got the release dates of the MBA Retina wrong. There seems to be plenty of evidence that Apple wanted to release the device for 2014 holiday sales. But what everybody (Apple, Intel, analysts, and sources) got wrong was the availability of the Intel chip to allow for fanless design that can drive retina without compromising processing power and battery life.

As a parenthetical matter, I will get the MacBook Air Retina regardless. But it is a bit squeamish to consider that the device will be using a relatively untested microprocessor (Broadwell U or Skylake). I hope Apple and Intel get this right.

My basis for disregarding the 9to5mac article lies not solely on the misprediction of the retina MBA launch (with the Broadwell fiasco I had already given up hope of seeing a new macbook-anything this year).

What my reason for disregarding the article is, is that the source stated:
"Finally, the WeiPhone.com user also notes that updated 13- and 15-inch MacBook Pros will be available in or around September. These updates will reportedly be mostly spec bumps rather than significant upgrades. The poster also mentioned that the iWatch isn’t imminent but would likely come late in the year."

He couldn't have been more wrong. The source predicted a refresh for MBP's and there wasn't even a mention of them this fall. Not only was he wrong about the timeline (he predicted in or around September). The whole thing never happened.

He's also wrong about the Apple watch, which will not be shipping until Q2 2015.

 

[rekhyt]

While Apple certainly could change the device quickly after introducing it (though I would prefer to believe previous mistakes have been learned from rather than are likely to be repeated), my point is rather that the delay in the initial introduction of the MacBook Air Retina is due to the fact that the Broadwell microprocessor won't do what those who have leaked details about this MBA iteration claim is being sought. Thus, Apple is arguably waiting on the availability of the Skylake microprocessor - which will provide true fanless and wireless charging design - before it can produce such a MacBook Air Retina.

Every report on the MBA Retina has assumed a Broadwell microprocessor. But if you research Broadwell and Skylake you find that only the latter is going to be able to fit the design that has been reported, which also has the advantage of explaining the delay in announcing the product.

Broadwell is a problematic and disappointing generation of microprocessor that Intel cannot wait to move beyond. But many of you who are cynical and bitter as to your Apple experiences seem to think the company will embrace such suspect technology even with the Skylake fix just around the corner. I am new to all of this but would like to believe Apple is better than that - and the facts as they are dribbling out seem to support my claim in this instance. I agree with Gildarts that Apple will bypass Broadwell and wait for Skylake.

Actually, Broadwell is the "tick", Skylake and Haswell are "tocks".
Broadwell is the successor of Haswell. Haswell was produced as a 22nm chip while Broadwell is being produced on 14nm.

Which means there were significant manufacturing changes to be made to switch production from 22nm to 14nm. Which in turn could explain why Intel has been experiencing difficulties with manufacturing Broadwell. Skylake will also be 14nm (cannonlake, the successor of Skylake will be 10nm). Intel has already stated that Skylake is on schedule and has proven to be unproblematic which is in stark contrast to Broadwell.


What Republius and I are trying to say is:
Given the advancements that will be made with Skylake. Is it plausible that Apple will be skipping Broadwell altogether?

I imagine they don't want an iPad 3 fiasco all over again. A lot of customers were upset that the iPad 4 came out soon after the iPad 3. The iPad 3 only had retina display compared to the iPad 2, while the iPad 4 brought significant performance improvements which were not implemented in the iPad 3. Also, iPad 3 has had the shortest product life of any iPad. They quickly took it off market. And at that time, iPad was still a relatively new product with a relatively small user base. Compare that to the established MacBook user base, and count in that mac's are 2-3 times the price of an ipad, the blowout could definitely be something for Apple to be considering.

So, with that in mind. Wouldn't it make sense for Apple to skip Broadwell altogether for their next MacBook refresh? I (and Republius) mean, why would Apple let Intel, who is basically late to their own schedule, mess up their product line? It just doesn't make sense to release a retina MBA on Broadwell in june when it's almost a given fact that the MBP's will get Skylake in the fall.

Edit: Also, Tim Cook stated in the earnings call that Apple Watch will not be shipping in Q1 2015. Which could mean that it will be shipping after WWDC, Q2. That could give them space to keep WWDC about Apple Watch and OS X & iOS and keep everything else for the fall.

Yep. I have to say I got burned by that.

According to my memory: The iPad 3 (the first iPad with Retina display) was released late Q1/early Q2. I bought the iPad during summer, and then swiftly afterwards (a little more than a month), the iPad 4 (with proper internals to drive the Retina display) was released.

I have to say I was really disappointed by that, and I have abstained from adopting the first generation of Retina devices ever since (dodged a bullet with the first generation of rMBPs).

When I bought my iPad 3, I bought it in the hope that it would be future-proof (Retina display is the biggest advancement for the iPad for a while now, in my own opinion). I didn't expect it to be obsolete in less than a year's time.

I'm really not sure what I'm going to do with the rMBA 12". On the one hand, I really need it (my current MBA 11" is getting old. It runs beautifully but I don't have enough storage (128 GB, unfortunately) or RAM (4 GB :/ ), but on the other hand, I don't want to buy into the first generation and then see the "proper" version of the laptop being released 6 months–1 year later, which can smoothly run the laptop.

:/

We've had integrated GPUs that can easily drive a retina display since Sandy Bridge, so that is certainly not a problem. I think the big hurdle for the retina MBA is the power consumption of the display. The rMBP had to increase the size of the battery by around 20% to adjust for it.

You sure about this?

I heard there were lagging issues that affected the first and second generation of the rMBP 15"s.

 

[motrek]

...
As a parenthetical matter, I will get the MacBook Air Retina regardless. But it is a bit squeamish to consider that the device will be using a relatively untested microprocessor (Broadwell U or Skylake). I hope Apple and Intel get this right.

How can they get it wrong? It's not like we haven't been through a million generations of Intel processors before, ticks and tocks and whatever else. Did you have the same trepidation about switching to Sandy Bridge, Ivy Bridge, Haswell, etc.?

Intel parts are basically the last things I worry about. Apple has had to do recalls on their laptops for graphics chips, faulty batteries, cracking top cases, and recently faulty SSDs. I can't remember the last time Apple had to recall a product due to a problem with an Intel part though.

 

[Serban]

The macbook air retina will have broadwell but with the iGPU better than HD5300.
And Macbook Pro updates will have skyline with thunderbolt 3

 

[Samtb]

Anyone think Apple will skip broadwell with the 15 inch and release it in a thinner form factor later this year?