Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Adam552

macrumors 6502
May 30, 2006
265
54
Liverpool, UK.
I would see your results for Dirt 3, Diablo 3, and Far Cry 3 and run tests on them as well on our multiple Macs.

Can you share your benchmarking procedures?

Using fraps, I'll share what I did to test each one when I've got the results.
Will be using Windows 7 with everything updated to latest version.
 

barefeats

macrumors 65816
Jul 6, 2000
1,058
19
Thank you folks for tests!

Please...

What about heat and noise on iMac 21.5 an 27?

Could be interesting to see also the differences from 680MX vs. 675MX.

I've been listening to music on headphones while testing so the noise is very pleasing -- ;-)

I'll monitor heat and noise when I run the Ray Trace 3D render in After Effects (GPU killer) and the Total Benchmark (CPU killer).

----------

Using fraps, I'll share what I did to test each one when I've got the results.

Yes. Quality settings would help -- especially if there are High or Ultra quality presets we can duplicate easily when we try running. Also, if you can run at 2560x1440 as well as 1920x1080 (or close to it), we can compare to the 21.5" iMac.
 

Ctrl2k

macrumors member
Oct 18, 2007
93
0
Barefeats - are you sure CUDA is enabled with Adobe suite? May have to force that on. Also can you include 2011 high end iMac for comparing? That's what I'm upgrading from purely for video production work.

Thanks to you and everyone for the tests! Mine comes Wed.
 

ottawaP5

macrumors newbie
Dec 14, 2012
7
0
Yes, I would like to know this as well, I did ask before and only got a vague reply. After playing WoW for a few minutes on a rMBP, you can clearly hear the fans spin up, and it is pretty audible. How does the 27" iMac with the 680mx compare ?

You definitely hear the fan as well. The sound from your game should drown it out since the fan noise comes out towards the rear.

As for heat, it is concentrated around the fan opening (beside the memory access panel). That part gets just as hot as the previous gen. The rest of the rear casing is just barely warm. Heat dissipation is much better then the previous gen.
 

kaellar

macrumors 6502
Nov 12, 2012
441
17
any news on OCing the 675MX/680MX atm?
since the temps are so low, I think it's time for those of you guys who have that GPUs to OC them.
 

Adam552

macrumors 6502
May 30, 2006
265
54
Liverpool, UK.
Yes. Quality settings would help -- especially if there are High or Ultra quality presets we can duplicate easily when we try running. Also, if you can run at 2560x1440 as well as 1920x1080 (or close to it), we can compare to the 21.5" iMac.

Yep was planning on testing High and Ultra with different AA settings or whatever lol at both resolutions. I can't tell which quality settings exactly at the moment as I have the games but I've never actually played them because my MBP has a 320M lol.
 

ottawaP5

macrumors newbie
Dec 14, 2012
7
0
I was wondering if someone could try Euro Truck Simulator 2 with Bootcamp? It doesn't seem to run all that well, there's a lot of intermittent stuttering and I don't even have the graphics maxed out.
 

racher

macrumors member
Apr 14, 2010
44
0
Seattle, WA
Barefeats - are you sure CUDA is enabled with Adobe suite? May have to force that on. Also can you include 2011 high end iMac for comparing? That's what I'm upgrading from purely for video production work.

Thanks to you and everyone for the tests! Mine comes Wed.

Ditto on the CUDA enabling. I'm upgrading from a 2010 MBP with Nvidia GT 330M card, and I'm sure the new iMac will blow that computer out of the water, but I'm curious on all computer comparisons. My fully loaded iMac will arrive Tuesday.

Thanks again for testing these pro apps for the benefit of (apparently) just a few of us!
 

James_C

macrumors 68030
Sep 13, 2002
2,817
1,822
Bristol, UK
You definitely hear the fan as well. The sound from your game should drown it out since the fan noise comes out towards the rear.

As for heat, it is concentrated around the fan opening (beside the memory access panel). That part gets just as hot as the previous gen. The rest of the rear casing is just barely warm. Heat dissipation is much better then the previous gen.

Well, not surprised, but disappointing. I was hoping that moving to a thinner form factor would not impact the ability iMac to run without significant fan noise, but it sounds like it is as bad as fans / cooling system on the new rMBP when put under load. Form > Function fail then as far as I am concerned.

The fans on my current Mac Pro don't spin up when I do something that stresses the GPU, and my Mac Pro is whisper quiet. Just daft that the iMac is so thin, so they can't have a quiet cooling system as well.
 
Last edited:

barefeats

macrumors 65816
Jul 6, 2000
1,058
19
Barefeats - are you sure CUDA is enabled with Adobe suite? May have to force that on. Also can you include 2011 high end iMac for comparing? That's what I'm upgrading from purely for video production work.

Thanks to you and everyone for the tests! Mine comes Wed.

Yes. I know. I edited the "cuda_supported_cards.txt" file in the Premiere Pro Contents folder to add the 680MX to the list. I've done any several times on the Mac Pro to make sure it "approved" GTX 570 and GTX 680 GPUs.

But I have to admit that some of the effects showed red when I thought the should show yellow. I'm wondering if I should list it as a 680M instead of a 680MX -- I'll try that to see if it changes anything -- and if it does, I'll post the updated numbers.
 
Last edited:

barefeats

macrumors 65816
Jul 6, 2000
1,058
19
RESULTS for GPU intensive Pro App functions:

Photoshop CS6 - render Iris Blur on 159M test image
2012 iMac 3.4 Core i7, GeForce GTX 680MX = 15.6 sec
2011 iMac 3.4 Core i7, Radeon HD 6970M = 19.1 sec
2012 iMac 2.9 Core i5, GeForce GT 650M = 25.5 sec
2012 Retina MBP 2.7 Core i7, GeForce GT 650M = 25.0 sec
2012 Mac mini 2.6 Core i7, Intel HD 4000 = 212.0 sec

Photoshop CS6 - render Tilt/Shift on 159M test image
2012 iMac 3.4 Core i7, GeForce GTX 680MX = 5.8 sec
2011 iMac 3.4 Core i7, Radeon HD 6970M = N/A
2012 iMac 2.9 Core i5, GeForce GT 650M = 8.5 sec
2012 Retina MBP 2.7 Core i7, GeForce GT 650M = 8.4 sec
2012 Mac mini 2.6 Core i7, Intel HD 4000 = 21.8 sec

Final Cut Pro X - render Directional Blur effect on 30 sec Apple ProRes 422 (HQ) 19x10 video
2012 iMac 3.4 Core i7, GeForce GTX 680MX = 10.0 sec
2011 iMac 3.4 i7, Radeon HD 6970M = 24.7 sec
2012 iMac 2.9 Core i5, GeForce GT 650M = 17.7 sec
2012 Retina MBP 2.7 Core i7, GeForce GT 650M = 13.4 sec
2012 Mac mini 2.6 Core i7, Intel HD 4000 = 32.2 sec

Final Cut Pro X - render Rain effect (lower is faster) on 30 sec Apple ProRes 422 (HQ) 19x10 video
2012 iMac 3.4 Core i7, GeForce GTX 680MX = 13.3 sec
2012 iMac 2.9 Core i5, GeForce GT 650M = 24.1 sec
2012 Mac mini 2.6 Core i7, Intel HD 4000 = 49.0 sec

Motion 5 - render RAM Preview of 600 frame HD template (higher is better)
2012 iMac 3.4 Core i7, GeForce GTX 680MX = 40.8 FPS
2011 iMac 3.4 Core i7, Radeon HD 6970M = 25.2 FPS
2012 iMac 2.9 Core i5, GeForce GT 650M = 17.0 FPS
2012 Retina MBP 2.7 Core i7, GeForce GT 650M = 19.9 FPS
2012 Mac mini 2.6 Core i7, Intel HD 4000 = 9.5 FPS
 

Ctrl2k

macrumors member
Oct 18, 2007
93
0
Yes. I know. I edited the "cuda_supported_cards.txt" file in the Premiere Pro Contents folder to add the 680MX to the list. I've done any several times on the Mac Pro to make sure it "approved" GTX 570 and GTX 680 GPUs.

But I have to admit that some of the effects showed red when I thought the should show yellow. I'm wondering if I should list it as a 680M instead of a 680MX -- I'll try that to see if it changes anything -- and if it does, I'll post the updated numbers.

Thank you sir. Impressive PS and FCPX numbers I see too.
 

Gnomeflyer

macrumors newbie
Oct 23, 2012
11
0
RESULTS for GPU intensive Pro App functions:

Photoshop CS6 - render Iris Blur on 159M test image
2012 iMac 3.4 Core i7, GeForce GTX 680MX = 15.6 sec
2011 iMac 3.4 Core i7, Radeon HD 6970M = 19.1 sec
2012 iMac 2.9 Core i5, GeForce GT 650M = 25.5 sec
2012 Retina MBP 2.7 Core i7, GeForce GT 650M = 25.0 sec
2012 Mac mini 2.6 Core i7, Intel HD 4000 = 212.0 sec

Photoshop CS6 - render Tilt/Shift on 159M test image
2012 iMac 3.4 Core i7, GeForce GTX 680MX = 5.8 sec
2011 iMac 3.4 Core i7, Radeon HD 6970M = N/A
2012 iMac 2.9 Core i5, GeForce GT 650M = 8.5 sec
2012 Retina MBP 2.7 Core i7, GeForce GT 650M = 8.4 sec
2012 Mac mini 2.6 Core i7, Intel HD 4000 = 21.8 sec

Final Cut Pro X - render Directional Blur effect on 30 sec Apple ProRes 422 (HQ) 19x10 video
2012 iMac 3.4 Core i7, GeForce GTX 680MX = 10.0 sec
2011 iMac 3.4 i7, Radeon HD 6970M = 24.7 sec
2012 iMac 2.9 Core i5, GeForce GT 650M = 17.7 sec
2012 Retina MBP 2.7 Core i7, GeForce GT 650M = 13.4 sec
2012 Mac mini 2.6 Core i7, Intel HD 4000 = 32.2 sec

Final Cut Pro X - render Rain effect (lower is faster) on 30 sec Apple ProRes 422 (HQ) 19x10 video
2012 iMac 3.4 Core i7, GeForce GTX 680MX = 13.3 sec
2012 iMac 2.9 Core i5, GeForce GT 650M = 24.1 sec
2012 Mac mini 2.6 Core i7, Intel HD 4000 = 49.0 sec

Motion 5 - render RAM Preview of 600 frame HD template (higher is better)
2012 iMac 3.4 Core i7, GeForce GTX 680MX = 40.8 FPS
2011 iMac 3.4 Core i7, Radeon HD 6970M = 25.2 FPS
2012 iMac 2.9 Core i5, GeForce GT 650M = 17.0 FPS
2012 Retina MBP 2.7 Core i7, GeForce GT 650M = 19.9 FPS
2012 Mac mini 2.6 Core i7, Intel HD 4000 = 9.5 FPS

Are these number being primarily alterted by the GPU, or does the CPU play a big part in them as well? Kind of a "green" question but I've never been a stats guy :confused:
 

SonomaFlyer

macrumors newbie
May 8, 2010
28
4
Apple apparently made the decision that most of the folks buying IMacs either won't push the limits of the unit's performance or won't notice/care about the noise.

Enthusiasts like us will benchmark em, push em and prod em to see what it can do... :rolleyes:
 

Blkant

macrumors regular
Jul 27, 2012
114
312
Apple apparently made the decision that most of the folks buying IMacs either won't push the limits of the unit's performance or won't notice/care about the noise.

Enthusiasts like us will benchmark em, push em and prod em to see what it can do... :rolleyes:

How do you make that? So far I'm hearing mostly I didn't notice or it wasn't to bad in regard to noise. Clearly previous macs it was a major issue, but so was heat and that seems to have been greatly improved. So hopefully once we get more thorough information we'll find that a similar attention has been given to the noise factor. After all, already having less heat should mean less need for noise.
 

barefeats

macrumors 65816
Jul 6, 2000
1,058
19
Are these number being primarily alterted by the GPU, or does the CPU play a big part in them as well? Kind of a "green" question but I've never been a stats guy :confused:

These particular functions of those pro apps are heavily GPU stressing and the CPU is idling along waiting for the GPU to finish. I confirmed this with OpenGL Driver Monitor with "CPU Wait for GPU" and "Current Free Video Memory" parameters.

There are many other functions that are CPU stressing. I plan to post those results on my website and possibly in another thread in this forum -- since this thread is about the 680MX GPU.
 

Jay9495

macrumors regular
Jul 11, 2011
125
0
Melbourne Australia
Far cry 3 results below

First off I'm using fraps, the iMac is not loud and I'm quite impressed regarding the results through bootcamp

Specs
i7 3.4Ghz
680mx
No fusion

Vsync is off

So at 1440p with everything on ultra, but no AA it averages at 20fps

At 1080p with everything on ultra no 2xAA it averages at 40fps, dipping to 30-35 at times

Tell me if you want me to modify the settings in any way
 
Last edited:

mattcube64

macrumors 65816
May 21, 2006
1,297
114
Missouri
Far cry 3 results below

First off I'm using fraps, the iMac is not loud and I'm quite impressed regarding the results through bootcamp

Specs
i7 3.4Ghz
680mx
No fusion

Vsync is off

So at 1440p with everything on ultra, but no AA it averages at 20fps

At 1080p with everything on ultra no 2xAA it averages at 40fps, dipping to 30-35 at times

Tell me if you want me to modify the settings in any way

Is the 1080p full-screen or are you running Window'd?
 

thekev

macrumors 604
Aug 5, 2010
7,005
3,343
RESULTS for GPU intensive Pro App functions:

Photoshop CS6 - render Iris Blur on 159M test image
2012 iMac 3.4 Core i7, GeForce GTX 680MX = 15.6 sec
2011 iMac 3.4 Core i7, Radeon HD 6970M = 19.1 sec
2012 iMac 2.9 Core i5, GeForce GT 650M = 25.5 sec
2012 Retina MBP 2.7 Core i7, GeForce GT 650M = 25.0 sec
2012 Mac mini 2.6 Core i7, Intel HD 4000 = 212.0 sec

Photoshop CS6 - render Tilt/Shift on 159M test image
2012 iMac 3.4 Core i7, GeForce GTX 680MX = 5.8 sec
2011 iMac 3.4 Core i7, Radeon HD 6970M = N/A
2012 iMac 2.9 Core i5, GeForce GT 650M = 8.5 sec
2012 Retina MBP 2.7 Core i7, GeForce GT 650M = 8.4 sec
2012 Mac mini 2.6 Core i7, Intel HD 4000 = 21.8 sec

Final Cut Pro X - render Directional Blur effect on 30 sec Apple ProRes 422 (HQ) 19x10 video
2012 iMac 3.4 Core i7, GeForce GTX 680MX = 10.0 sec
2011 iMac 3.4 i7, Radeon HD 6970M = 24.7 sec
2012 iMac 2.9 Core i5, GeForce GT 650M = 17.7 sec
2012 Retina MBP 2.7 Core i7, GeForce GT 650M = 13.4 sec
2012 Mac mini 2.6 Core i7, Intel HD 4000 = 32.2 sec

Final Cut Pro X - render Rain effect (lower is faster) on 30 sec Apple ProRes 422 (HQ) 19x10 video
2012 iMac 3.4 Core i7, GeForce GTX 680MX = 13.3 sec
2012 iMac 2.9 Core i5, GeForce GT 650M = 24.1 sec
2012 Mac mini 2.6 Core i7, Intel HD 4000 = 49.0 sec

Motion 5 - render RAM Preview of 600 frame HD template (higher is better)
2012 iMac 3.4 Core i7, GeForce GTX 680MX = 40.8 FPS
2011 iMac 3.4 Core i7, Radeon HD 6970M = 25.2 FPS
2012 iMac 2.9 Core i5, GeForce GT 650M = 17.0 FPS
2012 Retina MBP 2.7 Core i7, GeForce GT 650M = 19.9 FPS
2012 Mac mini 2.6 Core i7, Intel HD 4000 = 9.5 FPS

Is OpenCL still not working under OSX with the HD 4000, or is it just really slow? The 680MX looks quite nice there. On the Motion 5 test, only the imacs seem capable of previewing in real time.
 

MythicFrost

macrumors 68040
Mar 11, 2009
3,940
38
Australia
How do you make that? So far I'm hearing mostly I didn't notice or it wasn't to bad in regard to noise. Clearly previous macs it was a major issue, but so was heat and that seems to have been greatly improved. So hopefully once we get more thorough information we'll find that a similar attention has been given to the noise factor. After all, already having less heat should mean less need for noise.
Actually the 2011 models are pretty good in regard to noise. The fans in my base 2011 iMac really don't move from 1-1.2k RPM.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.