Looking at the benchmarks from the 6700HQ, having a new cpu architecture isn't doing anyone much good at the moment. What's the point of a new architecture if it performs just as good as the old one at the same wattage?
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Waiting for Skylake MBP thread
- Thread starter adildacoolset
- Start date
- Sort by reaction score
- Status
Haswell -> Skylake is a solid 10% clock for clock improvement with better power management. Its not huge, but its not nothing.
Evidently details don't matter to you either.
Keep firing the insults/comments and run away.
If you don't need another laptop right now, and your current laptop is performing, why not?
I´ve been waiting since March, I need a laptop and I want a Mac. However, I refuse to pay 1649 for old hardware. Yeah sure you can buy it cheaper somewhere else, but I´d like to buy it at Apple.
If they don´t announce new Macs in june though, I will start looking elsewhere or buy the 2015 model, if they give that a price drop.
If you don't need another laptop you shouldn't get another laptop in june either.
And obviously NOW you should wait. it's four weeks. But six months with a borrowed laptop, an underperforming laptop or even being without a laptop? If you think that's worth it for 10% performance gain you obviously don't use your laptop for work.
Being totally selfish, I have had to spend over £2500 of my savings this month, so if no MBP comes in June I can sit back and save a little longer without the temptation of buying a new MBP.
And to be fair I really don't think a worthy upgrade will come through, that along with the inflated price tag Apple will add to it, to persuade me to break the bank for it.
Time to look elsewhere before the WWDC dare I lol
And to be fair I really don't think a worthy upgrade will come through, that along with the inflated price tag Apple will add to it, to persuade me to break the bank for it.
Time to look elsewhere before the WWDC dare I lol
Last edited:
No it's not. It's like from minus ~5% to plus ~8% in various CPU workloads. I. e. on average it's much less than 10% compared to Broadwell.
However Skylake iGPU is much better than Broadwell - this is where most of the improvement went to IMHO. I'm not a fan of it, but to be fair there is an overall improvement after all.
Last edited:
I would also be interested in something like that. I think it is doable when they upgrade to skylake the leave it in the current 15" body. Should have plenty of space; here is wishing an hoping.
I do carry a fair bit of photography gear, Plus audio gear.
Not much going on as far as computing power. Waiting to see Iris Pro 580 benchmarks. Either way, TB3 should be a great addition.
Slight bump in compute power but I think the more interesting thing is we should see major bumps elsewhere. With the less power draw of Skylake and clock for clock it is faster then Hanswell leaves more power draw for a dgpu, beefier ram, faster SSD.
I understand the security concerns but it is true that machine learning algorithms will improve human life and simplify many petty tasks. At the same time I wonder how much privacy people really have nowdays, I think that we kid our selves when we say we have 'privacy' since a lot of people willingly give out personal information for the sake of likes. So does it really matter whether it be Apple's large servers or FB's? Also, assuming that most people don't have access to a computer that can parse through petabytes of data, let alone access the servers in the first place, I wouldn't mind having a data overload of essentially hiding the needle in a haystack. It's not impossible to find but it's improbable (no different to how it is now imho). So security vulnerability? no more than it is now... intrusive? depends on who i guess.
But yeah... I like the fact that Apple has the liberty to optimise its own software to run well on low specs (FCP on rMB for example) but it's unfortunate that the same doesn't apply to majority of other software one would use.
https://gfxbench.com/compare.jsp?be...+Pro+Graphics+580&D2=AMD+Radeon+R9+(TM)+M370X
Iris 580 pro is faster than the dGPU in the current 15" rMBP
Iris 580 pro is faster than the dGPU in the current 15" rMBP
And the difference is quite significant on a number of tasks!
Car chase, on screen: 3064 Frames (51.9 Fps) vs 1130 Frames (19.1 Fps)
Manhattan 3.1, on screen: 1949 Frames (31.4 Fps) vs 1293 Frames (20.9 Fps)
Manhattan, on screen: 2862 Frames (46.2 Fps) vs 1672 Frames (27.0 Fps)
Texturing, on screen: 23071 MTexels/s vs 16016 MTexels/s
And loose badly with 960m https://gfxbench.com/compare.jsp?be...)+Pro+Graphics+580&D2=NVIDIA+GeForce+GTX+960M and pascal has debut today (gtx1080 is 1.7x faster than predecesor gtx980 so nvidia can bring such improve in 1060m).
That image tells-the-tale on Skylake CPU performance. If you already have a model using the 4870HQ, like my 2015 rMBP 15", we're looking at less than 50 points difference in CPU performance, less than 1% improvement by going to Skylake. Granted, we'll probably see a higher-performing variant as a CTO option, but it'll be a trek to reach even a 10% improvement. That's just not where Intel has put their development time on their mobile chips, unfortunately.
Now, as others have noted, GPU performance should be much better (even just using the new Iris Pro 580 integrated chipset), and we should get a little more battery life (unless Apple reduces the battery size to maintain same battery life or similar).
Just need to have appropriate expectations, and I agree, anyone waiting out this new model strictly on CPU horsepower alone is going to be disappointed.
Last edited:
Please check before posting comparisons. Iris Pro was running under Windows using the OpenGL API while the 960m was running on Linux using the OpenGL Api. Once you switch the 960m to windows using the same OpenGl API the gap isn't abysmal, nor the 580 looses "badly". Add to that the fact that Apple has never included a dGPU with a TDP higher than a x50m. A x60m class dGPU has never been inside a rMBP nor it will, even more so now that the chasis is rumored to be thinner than a Macbook Air.
Last edited:
Comparision to 950m w/ Windows OpenGL API:
https://gfxbench.com/compare.jsp?be...&hwtype2=dGPU&hwname2=NVIDIA+GeForce+GTX+950M
Roughly equivalent in most benchmarks. Pretty impressive for an iGPU, but not particularly exciting.
True, the only thing that impresses me is the amount of performance they can extract using a significantly reduced power and area budget compared to Nvidia's and AMD's offerings. Intel never fails to impress me, on the GPU side of things. Imagine a full fledge dGPU from them. :O
Yes indeed, my fault with os version. https://gfxbench.com/compare.jsp?benchmark=gfx40&did1=32063337&os1=Windows&api1=gl&hwtype1=iGPU&hwname1=Intel(R)+Iris(TM)+Pro+Graphics+580&did2=25314219&os2=Windows&api2=gl&hwtype2=dGPU&hwname2=NVIDIA+GeForce+GTX+960M still lose badly (960m is 1.4x faster in Car Chase, 1.6x in Manhatan3.1, 1.52x in Manhatan, 1.2x in trex and probably its havy limited by cpu as it has over 300fps in this test). A x60m class was never in rMBP and that is a mistake as xps have it. Intel had never impressed me with they graphic gpu and won't this change this year (and we will later compare iris 580 to polaris and pascal, 1060m will be probably 2.5x faster).
I think what OSX7 is getting at is if you put these numbers into performance/watt, the 6770 is pretty darn impressive, and that certainly is in sync with Apple's apparent direction.
Hopefully we'll still see an Apple product with a Polaris/Pascal gpu at some point, though June might be a bit optimistic for that.
- Status