MacBook Air rev. c faster graphics?

[robeddie]

As someone in another thread noted, Apple is now claiming that the new Macbook Air has '6x faster' graphics compared to the original (rev.A) Macbook Air.

The rev. B page claimed only '4x faster' graphics.

This would of course, seem to imply that the graphics on the rev. C are 50% faster than on the rev. B.

Does anyone know if this is in fact true?

 

[mhnajjar]

http://www.apple.com/macbookair/graphics.html

No mention, so I don't think so!

 

[ayeying]

http://www.apple.com/macbookair/graphics.html

No mention, so I don't think so!

"Games get up to a 6x performance boost in MacBook Air, compared to the original model"

 

[glitch44]

"Games get up to a 6x performance boost in MacBook Air, compared to the original model"

It's the same chip as the Rev B, although the 9400m chip that it's in the Rev B and Rev C is faster than the original.

 

[mhnajjar]

"Games get up to a 6x performance boost in MacBook Air, compared to the original model"

Didn't see that lol

But what would make it clocked faster this time? Would there be a firmware for revB?

 

[mrossi]

This is a bit confusing... my rev.B w/ssd gets up to 80deg when doing anything even remotely intensive that involves both cores and the gfx chip. If the cpu and gpu hardware are the same same in rev.C, then can we assume Apple have changed/upgraded the cooling system? surely otherwise there will be big problems with the heat generated? (I haven't checked how the thermal paste is on mine yet but will do so soon...)

Didn't see that lol

But what would make it clocked faster this time? Would there be a firmware for revB?

I guess if the cooling system is the same in rev.C as rev.B then Apple obviously think it can cope with the extra heat. If this is the case, then I'd also really like to know how to overclock both the processor and gpu...

Anyone?

Edit: Ok, so maybe not all the hardware is the same...processor speed bump to 2.13GHz seems to correspond to the Intel SL9600 (as opposed to the 1.6GHz => SL9300 or 1.86GHz => SL9400, which were the choices for the Rev.B). But, if you look at this: http://ark.intel.com/Compare.aspx?ids=36689,37262, then they are certainly comparable (basically the same chip) so I think the argument about increased heat is a valid one.

 

[robeddie]

Of course we can all speculate ... but I guess it'll be a bit before we have any confirmation whether the video card in the rev. C is clocked faster than the card in the rev. B.

Another possibility is that performance is the same, but Apple did testing a little different and came up with 'new' numbers.

They have a history of giving somewhat 'suspect' performance numbers on their site, so that wouldn't surprise me.

 

[mrossi]

Of course we can all speculate ... but I guess it'll be a bit before we have any confirmation whether the video card in the rev. C is clocked faster than the card in the rev. B.

Another possibility is that performance is the same, but Apple did testing a little different and came up with 'new' numbers.

They have a history of giving somewhat 'suspect' performance numbers on their site, so that wouldn't surprise me.

I agree the numbers can be very suspect, especially when it comes to graphics cards, but taking a look at the XBench results (http://db.xbench.com/search.xhtml?text=macbook+air) it does seem as though both the processor and gpu clocks have been increased.

I'd _really_ like to know from someone who has purchased it whether or not the cooling hardware is the same. (Pity NC MacGuy didn't get one this time around as his disassembly post for the rev.B was incredibly useful https://forums.macrumors.com/threads/593227/)

 

[robeddie]

I agree the numbers can be very suspect, especially when it comes to graphics cards, but taking a look at the XBench results (http://db.xbench.com/search.xhtml?text=macbook+air) it does seem as though both the processor and gpu clocks have been increased.

Thanks for the link. I looked at the graphics specific tests of the new Airs on that chart and compared them to mine (a ver. b 1.86 ssd model).

The graphics numbers are NOWHERE NEAR 50% greater on the newest Airs, which Apple's web copy seems to imply (6x greater performance/4x greater performance).

The marginal difference I'm guessing is due to the greater cpu power (2.13 vs. 1.86)

So, based on that I'd have to say there is no difference in the clocking of the gpu between the 'fall 2008' Airs and the 'Summer 2009' Airs.

 

[ayeying]

I'd _really_ like to know from someone who has purchased it whether or not the cooling hardware is the same. (Pity NC MacGuy didn't get one this time around as his disassembly post for the rev.B was incredibly useful https://forums.macrumors.com/threads/593227/)

My system is on the truck right now. It should arrive very soon. I'll do a bottom panel removal and snap some pictures for you guys as soon as I can

 

[mrossi]

The graphics numbers are NOWHERE NEAR 50% greater on the newest Airs, which Apple's web copy seems to imply (6x greater performance/4x greater performance).

The marginal difference I'm guessing is due to the greater cpu power (2.13 vs. 1.86)

So, based on that I'd have to say there is no difference in the clocking of the gpu between the 'fall 2008' Airs and the 'Summer 2009' Airs.

I also looked at the rev.C XBench numbers a bit: Link 1 and Link 2

This is a quick comparison with my Rev.B w/ssd:

Test	Rev.C (1)	Rev.C (2)	Rev.C Avg	Rev.B	% Diff (C vs B)
CPU	143.34		143.53		143.44		121.6	15.22
Thread	257.15		281.91		269.53		259.79	3.61
Memory	176.56		177.36		176.96		166.23	6.06
Quartz	161.06		160.12		160.59		130.12	18.97
OpenGL	141.75		135.29		138.52		100.83	27.21
U.I.	240.34		271.16		255.75		155.1	39.35
Disk	84.72		86.09		85.41		80.08	6.23
					
Result	152.99		155.24		150.11		128.14	14.64

It's clearly not an increase on the scale that Apple are suggesting. Certainly not half as fast, but possibly 25%?

Does anyone know if the numbers for the gfx tests (Quartz/OpenGL) can be solely attributed to the increase in processor speed? Or must there also a bump in gpu speed?

Also, anyone care to explain the rather large (~40%) difference in the UI test? I'm not sure exactly what is being tested here...

My system is on the truck right now. It should arrive very soon. I'll do a bottom panel removal and snap some pictures for you guys as soon as I can

Awesome

 

[ayeying]

So a FedEx truck drove by... and drove away... this happened 3 months ago when I ordered this MacBook Air and Mac Pro... different trucks came with different packages. lol

I want my Rev C so I can snap some pix for you guys damn it

 

[dudeitsjay]

This is a bit confusing... my rev.B w/ssd gets up to 80deg when doing anything even remotely intensive that involves both cores and the gfx chip. If the cpu and gpu hardware are the same same in rev.C, then can we assume Apple have changed/upgraded the cooling system? surely otherwise there will be big problems with the heat generated? (I haven't checked how the thermal paste is on mine yet but will do so soon...)



I guess if the cooling system is the same in rev.C as rev.B then Apple obviously think it can cope with the extra heat. If this is the case, then I'd also really like to know how to overclock both the processor and gpu...

Anyone?

Edit: Ok, so maybe not all the hardware is the same...processor speed bump to 2.13GHz seems to correspond to the Intel SL9600 (as opposed to the 1.6GHz => SL9300 or 1.86GHz => SL9400, which were the choices for the Rev.B). But, if you look at this: http://ark.intel.com/Compare.aspx?ids=36689,37262, then they are certainly comparable (basically the same chip) so I think the argument about increased heat is a valid one.

Seeing the stripping of the C model, I noticed that they placed another layer of thermal shield/heatsink (depending on what they designed it to do; will need empirical experience to sort this one out) right onto the underbody... otherwise, there isn't much else i see different.

it did seem interesting the x4 better graphics turned to x6.1... i'm no expert but i'd say it'd have something to do with the improved cpu clockspeed which could somehow affect the integrated chip performance... and a 25% improvement is quite a lot these days... especially to go so underpublicized

 

[robeddie]

I also looked at the rev.C XBench numbers a bit: Link 1 and Link 2

This is a quick comparison with my Rev.B w/ssd:

Test	Rev.C (1)	Rev.C (2)	Rev.C Avg	Rev.B	% Diff (C vs B)
CPU	143.34		143.53		143.44		121.6	15.22
Thread	257.15		281.91		269.53		259.79	3.61
Memory	176.56		177.36		176.96		166.23	6.06
Quartz	161.06		160.12		160.59		130.12	18.97
OpenGL	141.75		135.29		138.52		100.83	27.21
U.I.	240.34		271.16		255.75		155.1	39.35
Disk	84.72		86.09		85.41		80.08	6.23
					
Result	152.99		155.24		150.11		128.14	14.64

It's clearly not an increase on the scale that Apple are suggesting. Certainly not half as fast, but possibly 25%?

Does anyone know if the numbers for the gfx tests (Quartz/OpenGL) can be solely attributed to the increase in processor speed? Or must there also a bump in gpu speed?

Also, anyone care to explain the rather large (~40%) difference in the UI test? I'm not sure exactly what is being tested here...



Awesome

On my 1.86 ssd rev. 2 Air... my Quartz results were close to yours (mine was about 133.

But my OpenGL result was a good bit better, about 115 (it varied between 116 and 114 during repeat tests)

So all in all, my graphics results are just a small percentage slower than the new top end rev. 3 (2.16ghz).