No, the 320M is not more powerful than the 8600M. Neither is the 330M 2x as powerful.

cathyy

macrumors 6502a
Original poster
Apr 12, 2008
727
1
So anyway, a bunch of people are using the notebookcheck.net 3DMark06 score of 3300 as the average score for the 8600M GT. Due to this, a lot of people are getting the wrong impression that the 320M is faster than the 8600M GT, and that the 330M is 2x more powerful.


There are several consistency issues which need to be addressed:

1) 8600M GT (DDR2) vs 8600M GT (GDDR3)
While they both have the same name, there are actually 2 different variants of the 8600M GT card. One with DDR2 memory and one with GDDR3 memory. (Note: this is NOT the same as system memory) The GDDR3 variant is between 20% - 30% faster than the DDR2 variant. Majority of the 8600Ms tested used DDR2 memory, while all the 8600Ms in Apple notebooks use GDDR3 memory.

2) 3DMark06 scores take the CPU into account
Your points are largely affected by how powerful your CPU is. Majority of the notebooks tested were also using the older Merom CPUs, and hence the average is lower. Just using a similar clocked Penryn CPU instead of a Merom CPU adds a few hundred more points.

3) Resolution difference
This is a very important factor. The standard testing resolution for 3DMark06 is 1280x1024, whereas the 3DMark06 test in this thread where the 320M scored 4700, was done at 1280x800. While 224 more vertical pixels doesn't sound like much, it actually makes a huge difference in points with benchmarks, possibly boosting the score by as much as 25%.


To back all this up further, we'll take a look at this old thread: Post your MBP Penryn 3DMark06 scores

If you read through this, the average 3DMark06 score for the Early 2008 MBP owners (8600M GT) is actually between 4200 - 4500. 3DMark06 tests on the 330M GT have been scoring close to 7000, which makes it around 60% better than the 8600M GT.

So far, the only 3DMark06 test on the 320M so far was done at 1280x800 so it's harder to compare, but yet it still gives us a rough gauge. Only one 8600M GT test was conducted at 1280x800 in that thread, and it scored 5500 points. This should place the 320M about 15% weaker than the 8600M GT.


I estimate this is roughly how powerful all the cards are relative to the 8600M GT, with an error of up to 5% per card:

9400M - 40%
320M - 85%
8600M GT - 100%
9600M GT - 110%
GT 330M - 160%
 

cathyy

macrumors 6502a
Original poster
Apr 12, 2008
727
1
Well, considering the nvidia 9400m gt was dog slow, the 320m is a killer gpu.
Yup, it's definitely a much better GPU than the 9400M. However it still seems to lose out to the ATI 5450 by a slight margin, which has a TDP of only 11W. I'm unsure of the TDP of the 320M, but the TDP of the 9400M was 12W. Arguably, both the 320M and 5450 are dog slow compared to today's cards. On the other hand, there is really no need for more graphics processing power for basic tasks, other than for Window's fancy Aero.

yup. why compare to the 8600gt? as long as it's better than the previous gen, that's all that really matters (to me anyway)
Read the opening paragraph. A lot of people were getting the wrong impression that the 320M is more powerful than the 8600M GT.
 
Comment

cathyy

macrumors 6502a
Original poster
Apr 12, 2008
727
1
>implying he told a cool story

as long as the overall performance is higher than the previous gen, I'll buy one.
In terms of graphics, the new MBP 13" blows away the old MBP 13". In terms of CPU processing power, it's exactly the same. However the old MBP 15" from Oct 2008 is still more powerful than the new MBP 13". Even the old SR 2007 MBPs are roughly as powerful as today's MBP 13".

If you owned any previous version of the MBP 13", there is pretty much no point 'upgrading' for this one because if the graphics card even mattered to you in the first place, you shouldn't be getting a MBP 13".


Yep another validation thread:D
Validation thread?

Don't be such an ass. This is exactly the sort of info I come to MacRumors for. Nice job, OP.
Thanks.


PS: I'm a girl.
 
Comment

believlle

macrumors newbie
Apr 15, 2010
24
2
actually i dont care wether 320M is better or worse than 8600M, they face to different user.

mbp 13 is just entry level, I think only 15 and 15 could be called "pro"
 
Comment

Pressure

macrumors 601
May 30, 2006
4,034
283
Denmark
The 15" MacBook Pro base model sees the biggest increase in performance.

Left 4 Dead (1440 x 900 - High Quality)
15-inch MacBook Pro (Late 2009) - GeForce 9400M
16.9 fps​
15-inch MacBook Pro (Mid 2010) - GeForce GT 330M
44.9 fps​

World of Warcraft (800 x 600 - High Quality)
15-inch MacBook Pro (Late 2009) - GeForce 9400M
19.1 fps​
15-inch MacBook Pro (Mid 2010) - GeForce GT 330M
52.3 fps​

Which may look impressive but the Radeon Mobility HD 5650 would have been heaps faster.
 
Comment

Puqq

macrumors regular
Oct 18, 2009
117
0
New 13" Graphics results: Pretty impressive, near 9600M GT

So, a couple of people are getting the wrong impression from this thread that the 320M is more powerful than the 8600M GT because it's scoring 4700 in 3DMark06 as compared to the average of 3300 scored by the 8600M GT at Notebookcheck.


3 very important things need to be mentioned:

1) 8600M GT (DDR2) vs 8600M GT (GDDR3)
While they both have the same name, there are actually 2 different variants of the 8600M GT card. One with DDR2 memory and one with GDDR3 memory. (Note: this is NOT the same as system memory) The GDDR3 variant is about 30% faster than the DDR2 variant. Majority of the 8600Ms tested used DDR2 memory, while all the 8600Ms in Apple notebooks use GDDR3 memory.

2) 3DMark06 scores take the CPU into account
Your points are largely affected by how powerful your CPU is. Majority of the notebooks tested were also using the older Merom CPUs, and hence the average is lower. Just using a similar clocked Penryn CPU instead of a Merom CPU adds a few hundred more points.

3) Resolution difference
This is probably the most important factor. The standard testing resolution for 3DMark06 is 1280x1024, whereas the 3DMark06 test in the thread stated was done at 1280x800. While 224 more vertical pixels doesn't sound like much, it actually makes a huge difference in points with benchmarks.


To back all this up further, we'll take a look at this old thread: Post your MBP Penryn 3DMark06 scores

These were the benchmarks done by Penryn MBP owners back in early 2008. On average, the users scored between 4200 - 4500 in 3DMark06, however all this was tested on 1280x1024 resolution. Only one person tested it at 1280x800, and he scored 5561 points. If the tests on the 320M were to be redone at 1280x1024 resolution, the average score will probably be lower than the scores we see in the thread above.

The 320M is a powerful card, but it's still weaker than the 8600M GT, and weaker than ATI's equivilant, the ATI 5450.
320M did 5561 in 3DMark06? Wow! I am not completely sure that this is right though

I would not expect 320M to be faster than 8600m GT, but VERY close indeed. Just because 310M is just marginally slower than 8600m GT and 320M has presumably better specs than 310M.

All in all, we should not rely on only one synthetic benchmark. A better comparison should come from tests in games.
 
Comment

Quetsche

macrumors member
Mar 5, 2010
80
0
The 15" MacBook Pro base model sees the biggest increase in performance.

Left 4 Dead (1440 x 900 - High Quality)
15-inch MacBook Pro (Late 2009) - GeForce 9400M
16.9 fps​
15-inch MacBook Pro (Mid 2010) - GeForce GT 330M
44.9 fps​

World of Warcraft (800 x 600 - High Quality)
15-inch MacBook Pro (Late 2009) - GeForce 9400M
19.1 fps​
15-inch MacBook Pro (Mid 2010) - GeForce GT 330M
52.3 fps​

Which may look impressive but the Radeon Mobility HD 5650 would have been heaps faster.
why would you compare a discrete GPU to an integrated one, and why would you compare the 9400M to the 330M in a thread about "320M vs 8600M" ?
:p
 
Comment

Pressure

macrumors 601
May 30, 2006
4,034
283
Denmark
why would you compare a discrete GPU to an integrated one, and why would you compare the 9400M to the 330M in a thread about "320M vs 8600M" ?
:p
Because the 2009 base model MacBook Pro had the Geforce 9400M and the new 2010 base model has the Geforce GT 330M.

It's a valid comparison.

Which is quite similar to the 2009 MacBook (Geforce 9400M) and the 2010 MacBook (Geforce 320M).

After all, both the Geforce 320M and the Geforce GT 330M has 48 "cores".
 
Comment

lamadude

macrumors 6502
Jan 12, 2006
432
0
Brussels, BE
Thanks for this! As an early 08 MBP owner considering switching to a new 13" MBP this was exactly the kind of thing I was wondering. I was expecting the 320m to be slower, but it seems the processor speed compensates this a little and the overall performance should be similar. For my limited gaming needs (mostly guild wars and some other old games) that should be enough :)
 
Comment

twitchtwice

macrumors member
Mar 19, 2010
81
0
this is interesting. i have a 8600mt GT pro and was wondering how the new 13" compared to it in testing. im going to wait to see how people run games like tf2 and css in bootcamp to see what kind of frames they are getting on the 13". that will make my decision
 
Comment

Cali3350

macrumors regular
Feb 16, 2009
249
0
It's actually entirely dependent on the game and setting. The 320m does have more processing power by a decent bit, so any shaded limited game (STALKER, Just Cause, Fallout) WILL be faster, and any bandwidth limited game (WOW, Doom 3) will be slower. The 320m is a surprisingly potent integrated GPU.
 
Comment

cathyy

macrumors 6502a
Original poster
Apr 12, 2008
727
1
320M did 5561 in 3DMark06? Wow! I am not completely sure that this is right though

I would not expect 320M to be faster than 8600m GT, but VERY close indeed. Just because 310M is just marginally slower than 8600m GT and 320M has presumably better specs than 310M.

All in all, we should not rely on only one synthetic benchmark. A better comparison should come from tests in games.
No, the 8600M GT did 5561 at 1280x800 in 3DMark06. The 320M did 4700 at 1280x800. The gap is definitely closing, which is not surprising because it has been 3 years already. However at the moment the 320M is still unable to pass the mid-range card from 3 years old.

Also, you're right that we shouldn't rely on synthetic benchmarks. They are not entirely accurate, but anyway it still gives a rough gauge on how well it performs overall.

Why make a new thread and not add this to the thread you are referring to ?
I did add this to the thread I was referring to, but many people weren't reading through the entire thread and missed it, and made the same assumption again that the 320M is better. I figured it'd be easier to just start a thread for it.

I'm not sure if this is relevant or not, but doesn't NVidia use "rebranding" of GPU's for Laptops?
The GT 330M is a rebrand of the GT 230M which is a rebrand of the GT 130M which is a rebrand of the 9650M GT which is a rebrand of the 8600M GT.
 
Comment

Pressure

macrumors 601
May 30, 2006
4,034
283
Denmark
The GT 330M is a rebrand of the GT 230M which is a rebrand of the GT 130M which is a rebrand of the 9650M GT which is a rebrand of the 8600M GT.
Actually, the Geforce GT 330M (GT216) is a rebrand of the Geforce GT 240/230M (GT216). And that is it.

The core configuration of the Geforce GT 240M (48:16:8, GT216) and the Geforce GT 130M (32:16:8, G96) are totally different.
 
Comment

Yixian

macrumors 65816
Jun 2, 2007
1,478
119
Europe
Actually, the Geforce GT 330M (GT216) is a rebrand of the Geforce GT 240/230M (GT216). And that is it.
Sorry no, the 330M GT is basically an 8600M GT, shrunk down with a few more shader cores.

It's ancient stuff really.
 
Comment
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.