Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Status
Not open for further replies.
I'm considering these alternatives to the 15" MBP:

  • Thinkpad P50
  • HP ZBook Studio G3
  • MSI WS60

On the Thinkpad and ZBook I can get a Skylake Xeon! I doubt this CPU be used in the new MBP.

I can also get matte displays (can't stand glossy). User replaceable parts. 32GB RAM (64GB on the P50). An Nvidia dGPU so I can run CUDA. OpenCL is a nice but it doesn't have support in machine learning. And they all have better cooling than the MBP.


The MSI WS60 is what the BOXX SLM is based on.

My workflow used to be MBP + Xserves, but with Apple discontinuing the Xserve I had to move to Linux servers. Switching now from my MBP to Linux will be much easier. Thanks Apple.

An anorexic laptop in this class is not important to me because I use two laptops: one powerful and one light (MBA). I would be so happy with a workstation class 15" MBP the size of the PowerBook G4 but that's not going to happen.
Thinkpad P50 does not and unfortunately will not have configurations with Intel Iris (Pro), yes it has dGPU but Iris (Pro) is not only the best iGPU, it also helps CPU perform better. Also it only has 1 Thunderbolt 3 port.
HP ZBook Studio is, from pure theoretical POV, the best spec possible. It's very small, allows Xeon with Intel Iris Pro, has 2 Thunderbolt 3 ports. However allegedly its fans are noisy as hell. :( I've not yet visited a brand shop to check this myself but unfortunately this info is throughout the internet.
Also consider Dell Precision 15 7510. Its weight is awful. I mean it's quite ok considering its power but comparing to rMBP or P50 or and even more - Zbook Studio - it's awful. But specs are good - Xeon with Iris Pro. Don't know about the noise.
 
  • Like
Reactions: 173080
http://www.notebookcheck.net/Intel-Skull-Canyon-NUC6i7KYK-Review.167509.0.html
"CineBench scores show the i7-6770HQ to be essentially identical to the i7-6700HQ." :D
I wonder if you also only read news titles and not the content? :D

Cinebench R15 CPU Multi 64Bit 670 vs 711 +6.12%
Cinebench R15 CPU Single 64Bit 145 vs 139 -4.14%
Cinebench R11.5 CPU Multi 64Bit 7.43 vs 7.76 +4.44%
Cinebench R11.5 CPU Single 64Bit 1.68 vs 1.66 -1.19%
Cinebench R10 Rendering Multiple CPUs 32Bit 20386 vs 21075 +3.38%
Cinebench R10 Rendering Single 32Bit 5364 vs 5536 +3.21%
wPrime 2.0x - 1024m 232.117s vs 225.673s +2.78%
Super Pi Mod 1.5 XS 32M 559.276s vs 538.979s +3.63%

However you should also consider that statistically its benchmark cannot be held as final truth - i7-6700HQ has 70+ benchmarks vs 1 benchmark of i7-6770HQ. And also thermal constraints of Skull Canyon and laptop might be quite different (they both allow TDP down configuration which is highly unlikely in gaming laptop but is quite likely in a NUC; also turbo mode can be sustained longer in gaming laptop).
[doublepost=1466019788][/doublepost]Holy... I haven't noticed literally THE NEXT sentences after the one you quoted.

CineBench scores show the i7-6770HQ to be essentially identical to the i7-6700HQ. This alone is impressive considering that the NUC is just a fraction of the size of gaming notebooks that carry the i7-6700HQ as an option.
The desktop-class i7-6700K is roughly 20 percent faster than the i7-6770HQ with its significantly higher TDP requirements (45 W vs. 91 W).
BTW 91 / 45 = +102% (vs +20% performance).

You could also look on AnandTech, quite different perspective.

[doublepost=1466019955][/doublepost]In other words, you might have said "you can't say this, we don't know yet!" and I would mostly agree. But your dumb smile as well as your stalking for all the posts mentioning Iris Pro on the last several pages and your misinterpreted quotes are a little off the chart.
 
Last edited:
I noticed, and still can't believe it, that no one already posted this.
So am I still waiting
[...]
Can't find a good reason
Can't find hope to believe in!
1840.png
1840.png
1840.png
1840.png
1840.png
1840.png
1840.png
1840.png
1840.png
1840.png
1840.png
1840.png
1840.png
1840.png


:D
 
Hmh, there's a shop nearby that sells the rMBP 13" 8gb 512gb ssd model for 1789€, is that a good price? The 16gb 256gb ssd model costs about the same too, would that be a better choice?
 
I wonder if you also only read news titles and not the content? :D

Cinebench R15 CPU Multi 64Bit 670 vs 711 +6.12%
Cinebench R15 CPU Single 64Bit 145 vs 139 -4.14%
Cinebench R11.5 CPU Multi 64Bit 7.43 vs 7.76 +4.44%
Cinebench R11.5 CPU Single 64Bit 1.68 vs 1.66 -1.19%
Cinebench R10 Rendering Multiple CPUs 32Bit 20386 vs 21075 +3.38%
Cinebench R10 Rendering Single 32Bit 5364 vs 5536 +3.21%
wPrime 2.0x - 1024m 232.117s vs 225.673s +2.78%
Super Pi Mod 1.5 XS 32M 559.276s vs 538.979s +3.63%
How much on avarage, +2,3% ? :D And of course thanks to iris 580 :D The 2% difference avarange in 8 test its impossible to feel. Yeah I comment many post about iris 580 because of so many foolishnes was said about this gpu in this topic ;) And your posts are in top of most supid one :) You said that thinkpad p50 is not so good because there is no iris 580. Why in the hell there should be iris if there is quadro 2000m(960m equivalent) that is like twice as fast?
 
Last edited:
You said that no thinkpad p50 is not so good because there is no iris 580. Why in the hell there should be iris if there is quadro 2000m(960m equivalent) that is like twice as fast?
Quote me please, it's not that far away. For both your statements.

I have not said any of my opinion, I've only provided factual information. Stop lying.
Thinkpad P50 does not and unfortunately will not have configurations with Intel Iris (Pro), yes it has dGPU but Iris (Pro) is not only the best iGPU, it also helps CPU perform better.

This is even funnier considering I'm thinking about buying Thinkpad P50 myself (and I think I've mentioned it in this exact thread months ago - you should know this considering you're watching all my posts).
Why in the hell there should be iris if there is quadro 2000m(960m equivalent) that is like twice as fast?
Maybe because I was specifically targeting CPU performance rather than GPU performance? You see, when 6700K, top desktop CPU has +102% TDP and only +20% performance, when there is no other way to get more performance per GHz with all other things the same on mobile except to buy Iris Pro CPU instead of non-Iris Pro CPU, +2-8% performance is quite big.
 
Last edited:
  • Like
Reactions: 173080
Well, thanks for nothing, Apple. If the new MBP comes out before my Air bites it(not likely) I'll probably buy it. Since it almost certainly will not, I'm going to be building a PC instead of buying a Pro and Thunderbolt display and it's probably going to save me a large amount of money.

Why anyone would pay full retail for 2 year old hardware is beyond me but I'm not falling for it.

It looks like my iPhone 6 will be my last Apple product.
 
  • Like
Reactions: j1104638
So indeed you don't read the content, you only read the titles. On "avarage" it's unknown.GFY. :)
"Thinkpad P50 does not and unfortunately will not have configurations with Intel Iris (Pro), yes it has dGPU but Iris (Pro) is not only the best iGPU, it also helps CPU perform better." - one of the most stupid sentence in whole this topic (and its quite big) so congratulations :D As I wrote why the hell there should be iris 580 if there is quadro2000m(equivalent of 960m) and iris doesn't improve cpu speed as benchmarks shows (2% on average in no difference at all, much more difference you can get just on different cooling system).
 
"Thinkpad P50 does not and unfortunately will not have configurations with Intel Iris (Pro), yes it has dGPU but Iris (Pro) is not only the best iGPU, it also helps CPU perform better." - one of the most stupid sentence in whole this topic (and its quite big) so congratulations :D
What you quoted is factual information. There is no opinion involved. I don't tell if this is good or bad, if this is price effective or not, if it is necessary or optional, if it's big or small, I only tell the facts that Thinkpad P50 spec sheet, contrary to the P70, says P50 will not offer Iris Pro CPU.
As I wrote why the hell there should be iris 580 if there is quadro2000m(equivalent of 960m) and iris doesn't improve cpu speed as benchmarks shows (2% on avarage in no difference at all, much more difference you can get just on different cooling system).
First, as thoroughly explained above, it's not "2% on avarage", it's unknown on average. But even if you consider it really to be +2%, then it's relatively quite big, considering +102% TDP gives only +20% performance and here +0% TDP gives +2% performance. There are no other ways to increase CPU performance in laptop for the same amount of GHz and all other things the same.
 
Last edited by a moderator:
"Thinkpad P50 does not and unfortunately will not have configurations with Intel Iris (Pro), yes it has dGPU but Iris (Pro) is not only the best iGPU, it also helps CPU perform better." - one of the most stupid sentence in whole this topic (and its quite big) so congratulations :D As I wrote why the hell there should be iris 580 if there is quadro2000m(equivalent of 960m) and iris doesn't improve cpu speed as benchmarks shows (2% on average in no difference at all, much more difference you can get just on different cooling system).
*sigh* It's not the GPU that improves speed on certain CPU bound workloads, it's the the massive LLC integrated with it. In this case a generous 128MB. You should head over to Anadtech they have a thorough analysis on this. There you can also find a 65W 5775C being equal to and sometimes surpassing a 95W 6700K. Again, the massive LLC helps in certain workloads. It should be even more relevant now that Skylake doesn't have to evict things from the L3 to store them in L4. That cumbersome step is gone. In pre-Skylake CPUs the L4 acted as a victim cache.
 
  • Like
Reactions: 173080 and j1104638
What you quoted is factual information. There is no opinion involved. I don't tell if this is good or bad, if this is price effective or not, if it is necessary or optional, if it's big or small, I only tell the facts that Thinkpad P50 spec sheet, contrary to the P70, says P50 will not offer Iris Pro CPU. First, as thoroughly explained above, it's not "2% on avarage", it's unknown on average. But even if you consider it really to be +2%, then it's relatively quite big, considering +102% TDP gives only +20% performance and here +0% TDP gives +2% performance. There are no other ways to increase CPU performance in laptop for the same amount of GHz and all other things the same.
Lol :eek: +2% is quite big :D Eot :D
[doublepost=1466026560][/doublepost]
*sigh* It's not the GPU that improves speed on certain CPU bound workloads, it's the the massive LLC integrated with it. In this case a generous 128MB. You should head over to Anadtech they have a thorough analysis on this. There you can also find a 65W 5775C being equal to and sometimes surpassing a 95W 6700K.
Cool but l4 cache is not used by cpu (anandtech says it can be used theoretically but there is no benchmark or real situtatin that proof this). Notebookcheck skull canon test shows that 6770hq is practically identicall in cpu performance as 6700hq
 
Last edited by a moderator:
Lol :eek: +2% is quite big :D Eot :D
[doublepost=1466026560][/doublepost]
Cool but l4 cache is not used by cpu (anandtech says it can be used theoretically but there is no benchmark or real situtatin that proof this). Notebookcheck skull canon test shows that 6770hq is practically identicall in cpu performance as 6700hq
Because as you said it's not currently used by the CPU, and hence why I said certain workloads. Keeping something in cache is faster than retrieving it from RAM. I doubt a benchmark would expose or take advantage of that LLC.
 
  • Like
Reactions: 173080 and j1104638
Because as you said it's not currently used by the CPU, and hence why I said certain workloads. Keeping something in cache is faster than retrieving it from RAM. I doubt a benchmark would expose or take advantage of that LLC.
Because you want this to be true :) No benefits in real life situation.
 
+2% is quite big
You've finally understood, good.
Cool but l4 cache is not used by cpu (anandtech says it can be used theoretically but there is no benchmark or real situtatin that proof this). Notebookcheck skull canon test shows that 6770hq is practically identicall in cpu performance as 6700hq
Reread AnandTech please. There is a certain difference between Haswell Iris Pros and Skylake (and, if I remember right, Broadwell) Iris and Iris Pros. In layman terms Haswell required applications to be specifically altered in order to use L4 cache while for Skylake it should be used transparently without such alterations.

Also, as @OSX7 rightly mentioned, look at 5775C vs 6700K benchmarks which were executed in similar TDP environment constraints contrary to the Skull Canyon vs Gaming Laptops.
 
Because you want this to be true :) No benefits in real life situation.
Certainly, the scenarios would be very limited, I'm not disagreeing on this. But to discard it outright? What you said would apply to you: "Because you want this to be true :)"
 
  • Like
Reactions: j1104638
Because you want this to be true :) No benefits in real life situation.
Wrong again. And we're not summoning @WRONG this time.
[doublepost=1466027238][/doublepost]
Certainly, the scenarios would be very limited, I'm not disagreeing on this. But to discard it outright? What you said would apply to you: "Because you want this to be true :)"
+1. Scenarios will be limited because not all the scenarios are cache-bound.
 
  • Like
Reactions: OSX7
What does 'Q4' mean in terms of Apple's shipping schedule?

Rereading rumors I forgot that one recent report said the new MBP's will be announced in June and begin shipping in August. That would be awesome.

Maybe 13" reveal in roughly 30 days or so, and then the 15" is announced alongside the iPhone/watch event in September?
 
Status
Not open for further replies.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.