Waiting for Haswell MBP Mega Thread

[blueTattoo]

Note that the PCH (formerly south bridge, with a little bit of north bridge functionality added) is on the chip with the CPU. The next step (which I speculate will come two years from now with Skylake) will be to integrate them into one die.

Nope, that's L4 cache (128MB). PCH (LinxPoint LP) is on the same package with cpu on ULV chips only.

 

[john123]

With a custom EFI firmware and/or Mac OS X build, Apple could perfectly disable/hide the dGPU to GeekBench application. We cannot draw any conclusion yet. The probability is still 0,5...

I'm not trying to be obnoxious, but just because uncertainty exists doesn't make p=0.5. The probability that this is what happened is extremely low because throughout many iterations of different dGPUs, Geekbench has consistently detected them. It's not really a coin flip here.

I disagree. Even Apple's marketing would have a hard time to explain the drop of 24% (in some cases 60-80%) in graphics performance. And the better battery life is pure speculation. The extra space from the missing dedicated card could be used for thinness. The power consumption from the Iris 5200 is not that much lower than the combination of the 4600+750m: http://www.computerbase.de/artikel/grafikkarten/2013/intel-iris-pro-5200-grafik-im-test/6/

I don't think that the expected trade-off lower performance / better battery life would help sales, because I think there a lot of users out there, that are interested in specs and performance. 15MBPR is the last portable mac with good graphic performance. For some users this is the main reason to pay the extra 600$ in comparison to the 13MBPR. These people will not bring extra cash to Apple anymore, because the base 13 MBPR + 800$ Windows PC/ Notebook would be the better solution for them. Futhermore the product differentiation between 13 MBPR and 15MBPR would be difficult. Why spend more cash for the same compromises?

Honestly, I've done marketing for most of the last ten years. You're confusing the objective truth with what marketing people do, which is spin. If you need an example of the Apple spin machine in overdrive, think back to the Megahertz Myth. And people actually believed it.

It doesn't matter whether the Iris 5200 actually is X% slower than the 650M on average because there is no counter-marketing organization out there telling consumers that. All Apple needs is a couple benchmarks in specific tasks that show an improvement and generate a pretty bar graph. Anandtech's own review demonstrates this is certainly possible. A likely target will be a compute task operation of some sort.

As for product differentiation, again the marketing collateral doesn't agree. If you look at how the 15" is marketed now, the processor, design, screen/pixels, and graphics all share air time. The dGPU is hardly the differentiator. In fact–and here's the kicker–the graphics card name and the name "NVIDIA" don't even appear on Apple's MBP "performance" page at all except as a footnote. That's pretty damning.

In short, I think you're confusing what you want and the objective truth with what's likely in a world with spin. The segment of the market that is constituted by people like you and me who know the objective truth is actually quite small, relatively speaking. And even then, most people in this thread will whine and moan and still buy the new models.

 

[jafingi]

I'm not trying to be obnoxious, but just because uncertainty exists doesn't make p=0.5. The probability that this is what happened is extremely low because throughout many iterations of different dGPUs, Geekbench has consistently detected them. It's not really a coin flip here.



Honestly, I've done marketing for most of the last ten years. You're confusing the objective truth with what marketing people do, which is spin. If you need an example of the Apple spin machine in overdrive, think back to the Megahertz Myth. And people actually believed it.

It doesn't matter whether the Iris 5200 actually is X% slower than the 650M on average because there is no counter-marketing organization out there telling consumers that. All Apple needs is a couple benchmarks in specific tasks that show an improvement and generate a pretty bar graph. Anandtech's own review demonstrates this is certainly possible. A likely target will be a compute task operation of some sort.

As for product differentiation, again the marketing collateral doesn't agree. If you look at how the 15" is marketed now, the processor, design, screen/pixels, and graphics all share air time. The dGPU is hardly the differentiator. In fact–and here's the kicker–the graphics card name and the name "NVIDIA" don't even appear on Apple's MBP "performance" page at all except as a footnote. That's pretty damning.

In short, I think you're confusing what you want and the objective truth with what's likely in a world with spin. The segment of the market that is constituted by people like you and me who know the objective truth is actually quite small, relatively speaking. And even then, most people in this thread will whine and moan and still buy the new models.

So true!

If you see Apple's marketing material, they don't mention Nvidia GeForce as big features. Only if you look at the raw specs or a little section, you can see it is Nvidia.

Fact is, that most of the people who buy MBP's don't look at this, and don't understand it. They just need a laptop that turns on, has a nice screen, and works. It's only us geeks/pros who actually look at the raw specs.

For Apple, marketing wise, there will be no difference by using Iris Pro vs. GeForce. The marketing material would be the same; however, with Iris Pro, they could easily market it as more energy efficient = better battery life + take some benchmarks where it performs better than the 2012/2013 model.

Again, Apple won't ever do a performance where it writes "Iris Pro vs. GeForce 650M" - they'll write "Late 2013 vs. Late 2012" etc.

Also, they won't show us the raw FPS/score results, but instead use the classic "2.5x better" labels etc.

So I think there definitely is a possibility for Apple to only use Iris Pro. Marketing-wise they would only benefit from that, as they can market the rMBP as having better battery life + better performance.

And that Iris Pro closeup just looks sooo nice!

 

[mcarling]

Nope, that's L4 cache (128MB). PCH (LinxPoint LP) is on the same package with cpu on ULV chips only.

Thanks for the correction. I still expect Intel to integrate the PCH on die with the CPU within the next few years.

 

[cs02rm0]

This thread feels like just staring at this for months.

Waiting, waiting, still waiting...

 

[macshoair]

First, I wanna tell people who were trying to speculate from the newly added rMBP (Feb 2013) in the stores with 30 business days shipping time, that they were totall wrong. The shipping time now is 1-3 business days

Second, I have a question to everyone who is waiting very patiently for the release of the Haswell version in October and please tolerate my question for the sake of discussion. What would you do if October (no, let's say November) is gone and Apple did not announce any new rMBP with Hasweel, what would you do, would you continue to wait?

----------

Oh, and I am just saying, who would buy the base rMBP 15" (Feb 2013) as refurbished for $1,869.00 if I can get it new for $1,900 from Best Buy (education discount and Apple TV) or from Apple (with education discount and iTunes Gard)?!!!!!!!!!

 

[Quu]

Second, I have a question to everyone who is waiting very patiently for the release of the Haswell version in October and please tolerate my question for the sake of discussion. What would you do if October (no, let's say November) is gone and Apple did not announce any new rMBP with Hasweel, what would you do, would you continue to wait?

I would continue to wait. I've waited this long.

 

[macshoair]

I would continue to wait. I've waited this long.

Well, I say that this confirms the fact that some people suggested that we just want it but do not need it. We are in a great luxury, I guess...

 

[Vanilla35]

What would you do if October (no, let's say November) is gone and Apple did not announce any new rMBP with Hasweel, what would you do, would you continue to wait?

I would wait because this is not my first laptop, so I'd just wait it out. To me, it would be obvious to wait even longer because you've already waited this long. But for those who are less patient I'm sure they'd say screw it. Only thing I would care about in has well is a chance at a better dgpu. If it comes with only iris pro, depending on how reviews/benchmarks are on it would determine whether I got the new one, or save a few dollars for getting the old, dgpu

 

[Narcaz]

Maybe I have a european bias. Apple advertises the name Geforce, shows benchmarks and the possibility to play games here.

I agree with you, that the marketing could create some nice bars and say that the 5200 is the best solution for the 15 MBPR. But from a hardware performance perspective it is not.

And the better battery life is speculation from you. Only Apple know if is getting better or not. The power consumption is not really better and they tend to make devices thinner. So we could end with a thinner, less powerful device than last year with the same battery life.

 

[adjeff8]

Its not that simple.
Cos for what your suggesting to work, you need a full re design of OSX.
If they would just add touch screen now, it would be very awkward to use, because the OS is not optimized for it.

Win 8 is the other way around. Its been optimized for touch, so all who don't have it whines cos its much more user unfriendly, if you don't have touch.

So if Apple where moving to touch on their laptops, we would have seen big changes in OSX before hand.


Also if they moved to touch, their stationary computers would have to be changed to touch also, and be even more awkward to use then the laptops.

If you have tried touch on a laptop, you quickly find out its very uncomfortable. Sure you can do small thing like scrolling with just your thumb, like some one said.
But that is far from all you do. Having to move your whole arm to press things on a near vertical screen, is not good ergonomically.
And just imagine having to use it on a 25-35" screen. Just to press things on different sides of the screen would be agonizing.

Now some would say: Yes but you could use both mouse and touch.
That is true. But then you have to use either one hand on the screen, and the other on the mouse. Or move your hand between the mouse and the screen.
In the end its just much more simple and ergonomic, to not use the touch.

The only realm where a touch screen could work, is on a maximum of maybe a 11" laptop. But even there, it would be easier to use the mouse.
And optimizing OSX for a 11" would be bonkers.

A tablet is a totally different story, cos you hold the tablet with 1 or 2 hands.
And the iOS is optimized for touch.


Conclusion: Apple moving to touch with computers looking like they do today, would be one of the worst moves they can do.

I hear and understand you. I can see it being a pain for extensive use. But when people like you say "It would be the worst thing that they can do" isn't that just a bit hyperbolic when all you would have to do if you didn't like it is just not use it?

 

[thundersteele]

First, I wanna tell people who were trying to speculate from the newly added rMBP (Feb 2013) in the stores with 30 business days shipping time, that they were totall wrong. The shipping time now is 1-3 business days

Damn... good job at killing dreams

Well, I say that this confirms the fact that some people suggested that we just want it but do not need it. We are in a great luxury, I guess...

Of course... most of us don't need it, in the sense that we will continue to live and study or earn money even if we don't get a new Haswell MBP.
We want it, or at least we want to know what the specs are so that we can tell everybody why we are not going to buy it

 

[adjeff8]

Huh? That's not only contradictory but clear as mud.

What do you think the "Back to the Mac" campaign is all about? OSX is now riddled with the same iOS "features" you find on the iPad. So no, Apple is not about letting their devices be true to "themselves". It's just the opposite.

The last OS X to be pretty much independent from Apple's silly-assed "be-all-things-to-all-people" integration was Snow Leopard (IMO).

I'm one of those Mac users that's all for the integration of OS X and iOS, but Krazy B is right, that statement was contradictory and clear as mud

----------

I actually like Mission Control better than Spaces. I used to have every single application tied to a specific space, but now I have some freedom and it's actually pretty liberating. Plus the four-finger-swipe between workspaces is awesome and well worth not having vertical spaces IMO. Obviously not as awesome if you have a desktop/mouse set up though.

Full screen apps on the other hand, I don't understand either. I've tried using them, but I don't see any advantage to simply maximizing the window completely. I like having my menubar and having my dock pop up instantly. And sharing the desktop with other windows.

Ima have to agree with you. I used to use Spaces religiously. When they went to mission control I couldn't understand why they had to redesign a perfect thing. But I've gotten used to it and now when I work on my step mothers computer, which still has Spaces, I clearly see how Mission control with its 4 finger swipe is more convenient.

 

[Vanilla35]

I'm not trying to be obnoxious, but just because uncertainty exists doesn't make p=0.5. The probability that this is what happened is extremely low because throughout many iterations of different dGPUs, Geekbench has consistently detected them. It's not really a coin flip here.



Honestly, I've done marketing for most of the last ten years. You're confusing the objective truth with what marketing people do, which is spin. If you need an example of the Apple spin machine in overdrive, think back to the Megahertz Myth. And people actually believed it.

It doesn't matter whether the Iris 5200 actually is X% slower than the 650M on average because there is no counter-marketing organization out there telling consumers that. All Apple needs is a couple benchmarks in specific tasks that show an improvement and generate a pretty bar graph. Anandtech's own review demonstrates this is certainly possible. A likely target will be a compute task operation of some sort.

As for product differentiation, again the marketing collateral doesn't agree. If you look at how the 15" is marketed now, the processor, design, screen/pixels, and graphics all share air time. The dGPU is hardly the differentiator. In fact–and here's the kicker–the graphics card name and the name "NVIDIA" don't even appear on Apple's MBP "performance" page at all except as a footnote. That's pretty damning.

In short, I think you're confusing what you want and the objective truth with what's likely in a world with spin. The segment of the market that is constituted by people like you and me who know the objective truth is actually quite small, relatively speaking. And even then, most people in this thread will whine and moan and still buy the new models.

It'll be a sad day when apple takes the dGPU out of their Pro line, and replaces it with less powerful integrated graphics. It's even more sad that I don't have any doubt they'll do this. I'm just hoping that everything we've heard of so far is a misappropriation, but I'm even hard pressed to be that naive.

 

[GSPice]

Anyone have updated info on the possibility of 2x16GB SODIMMs in the Haswell MBPr?

 

[jafingi]

Anyone have updated info on the possibility of 2x16GB SODIMMs in the Haswell MBPr?

No, there are not SODIMMS in the rMBP, and there never has been...

Apple uses proprietary blocks RAM that are soldered on. Not SODIMM.

----------

Second, I have a question to everyone who is waiting very patiently for the release of the Haswell version in October and please tolerate my question for the sake of discussion. What would you do if October (no, let's say November) is gone and Apple did not announce any new rMBP with Hasweel, what would you do, would you continue to wait?

Of course I would wait. I'm not in serious need of a new laptop, and when I'm paying $3500 for a new laptop, I want the latest and greatest. Not a year old (and last generation) CPU. Haswell might not be the biggest performance jump, but I'm not going to use that much money on an Ivy Bridge laptop that I'm going to use the next 4-5 years, when I'm not in a hurry.

 

[Quu]

Well, I say that this confirms the fact that some people suggested that we just want it but do not need it. We are in a great luxury, I guess...

My current MacBook Pro is 4 years old. It will be 5 years old in February 2014.

I want to replace it. But I'm not going to replace it with a computer that is already a year out of date when we are so close to a Haswell update with better Wireless and Storage.

The rubber grommets have fallen off my Notebook, its USB ports are worn and one of them doesn't work reliably any more. My trackpad and keyboard sometimes stop working for a few seconds. I get 1 hour of Battery Life now instead of the 8 Hours (!) it came with. But it works, it works enough that I can wait for the Retina MacBook Pro update.

 

[GSPice]

No, there are not SODIMMS in the rMBP, and there never has been...

Apple uses proprietary blocks RAM that are soldered on. Not SODIMM.

Sorry, I knew that the ram is soldered. Let me rephrase. What are the chances of 32GB ram being an option in the mbpr.

 

[bill-p]

Sorry, I knew that the ram is soldered. Let me rephrase. What are the chances of 32GB ram being an option in the mbpr.

There's no 2x 16GB RAM modules on the market right now.

So the chances that the next rMBP will have that? Pretty much zero. Unless Apple pulls a miracle, but I suspect they can't because they don't manufacture their own RAM modules.

 

[jafingi]

Sorry, I knew that the ram is soldered. Let me rephrase. What are the chances of 32GB ram being an option in the mbpr.

I don't think the chances are very big.

Apple uses 2GB Hynix H5TC2G83CFR modules in the current rMBP (8GB) and 4GB modules (H5TC4G83MFR) in the 16GB.

They have 8GB modules (H5TC8G83MMR), which could be used for a 32GB version. However, I don't find this likely. It is a relatively new package (revision 1.1 was released in April 2013).

 

[FullerJ]

If Apple decided to pull some crap like that I would have no choice but purchase the current rMBP. I'm hoping and praying the haswell rMBP are released in mid-september/early-october because I start school mid september. I've held off buying the MBA in hopes that Apple smartens up and releases an updated rMBP before the back to school season is over.

 

[mneblett]

I've held off buying the MBA in hopes that Apple smartens up ...

Do you really think Apple isn't smart enough to figure out when/how to schedule production/marketing/product release to meet its objectives???

Granted, they may not be your objectives, but given their ability to create a legion of loyal followers (second only to Harley-Davison's ability to sell the "life style") it doesn't appear that there's much "smarten[ing] up" for them to do.

 

[FullerJ]

Do you really think Apple isn't smart enough to figure out when/how to schedule production/marketing/product release to meet its objectives???

Granted, they may not be your objectives, but given their ability to create a legion of loyal followers (second only to Harley-Davison's ability to sell the "life style") it doesn't appear that there's much "smarten[ing] up" for them to do.

You're absolutely right. Look at the rMBP last year. IT FAILED HARD. Releasing a new line of computers in October? Yeah because thats when parents want to buy their college student a laptop...halfway into the semester.

 

[zOlid]

I hear and understand you. I can see it being a pain for extensive use. But when people like you say "It would be the worst thing that they can do" isn't that just a bit hyperbolic when all you would have to do if you didn't like it is just not use it?

Absolutely not.
Cos that's not the issue.
Its a either you move the OS in that direction or you don't.
A touch optimized OS is a pain to use with out touch. Or at the very least much more annoying then a OS optimized for a mouse.
And its the same the other way around.

So if Apple would move for a touch bases OSX it would make the OS much worse to use if you don't use touch.

Just look at windows 8 as an example. Their you have both that's true.
But they are intertwined. Its easier to use a touch screen on some parts and mouse on others. They want to move to touch more and more. But older programs and computers don't allow it. So your stuck with a 2 based system.
Where if you use Metro its very optimized for touch. But way more annoying to use if you only use a mouse. And that is the biggest reason Windows 8 is hated by many.


The only way i see a touch based OS working is with a iPad (or such) on the side that is showing some controls or buttons that you can touch with your hand. Now this would be a pain on a laptop i guess. Or maybe a touch based screen, where the mouse pad is now but bigger.
That is one way you could go i guess but it would take some serious magic for it to work good.

There are apps and programs that utilize the iPad for some controls and shortcuts. Instead of the classic keyboard shortcuts. I've heard there is one for Photoshop (i think) that is apparently the best thing since steak sauce.

 

[tmiw]

You're absolutely right. Look at the rMBP last year. IT FAILED HARD. Releasing a new line of computers in October? Yeah because thats when parents want to buy their college student a laptop...halfway into the semester.

I think this is sarcasm, but if it isn't: Apple wants students to buy the MBA, not the rMBP. And that came out before BTS.