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

Results of 2011 iMac Overclock (bootcamp, prelim)

iSayuSay said:
You better elaborate all your work here and update it on your first post. You can copy other member OC result too with identical GPU.

GameTitle/MemClock/CoreClock/FPS/GPUTemp ... something like that. I appreciate your work, just don´t make your result sporadic along with your posts, this thread could be awesome like another thread about OC last year model 5850M
Click to expand...

Good tip. That's why I was getting extra annoyed at all the side banter. I'm hoping to make this thread just that good. Didn't think about editing my 1st post :eek:
 
5piN said:
Could you run the three Just Cause 2 benchmarks and post results with settings and then I'll run the same on mine, just want to see if the extra 1gb on your 6970 makes any difference. .... oh yea on stock settings please.

Thanks in advance.
Click to expand...

Sure. I doubt it will make a big difference. But we'll see!
 
swarleystinson said:
IS IT POSSIBLE TO POST ONE THING ABOUT OVERCLOCKING WITHOUT HAVING SOMEONE POST:
1) It's dangerous
2) You should've bought a gaming rig.


I don't have room for 2 computers in my apartment. I chose to go with iMac because it's the best all-in-one available. Furthermore, Windows 7, though leaps and bounds better than vista and all-around a great OS, I prefer Mac OSX. I use a great deal of applications that are unique to Mac and have my content geared towards Mac use.

NOW PLEASE STOP TROLLING!

For those who are actually interested in discussing overclocking, please see next post...
Click to expand...

Fair enough mate. Good luck with your OCing, I am sure other people who game on their iMacs will be interested in this. I never looked I not this but is there a possibility to upgrade the fans in the iMac. Should be quieter and more efficient
 
You mentioned earlier that you did a benchmark on Just Cause 2 and I was hoping you might be willing to do a few tests for me, but if not, no worries.

With stock clocks, could you run the "Dark Tower" benchmark with these settings: http://yfrog.com/5g2011052400001j

And then once more but with 8x AA and then let me know the results of both.

Also thanks for posting these OC'd results! :) Cheers
 
MythicFrost said:
You mentioned earlier that you did a benchmark on Just Cause 2 and I was hoping you might be willing to do a few tests for me, but if not, no worries.

With stock clocks, could you run the "Dark Tower" benchmark with these settings: http://yfrog.com/5g2011052400001j

And then once more but with 8x AA and then let me know the results of both.

Also thanks for posting these OC'd results! :) Cheers
Click to expand...

Hey .... i had some time today so i ran the benchmark for you ..albeit i have the 1gb version
 

Attachments

  • noaa.PNG
    noaa.PNG
    835 KB · Views: 143
  • 8xaa.PNG
    8xaa.PNG
    889 KB · Views: 114
Just Cause 2 Dark Tower Benchmark results (same settings as above):

680/900:
AA off - 29.34 Avg Fps
8x AA - 21.74 Avg Fps

800/1100:
AA off - 34.35 Avg Fps
8x AA - 25.81 Avg Fps
 
Now your first post look beautiful :p

I´m still waiting my iMac to be arrived, i´ll share some result here once it come :)

Anyway all your FPS result was recorded using FRAPS? So we could expect higher FPS when FRAPS turned off, sir?
 
Thanks a lot to both of you for running those benchmarks! That's a pretty good improvement from the overclock, too.

I wonder how Metro 2033 will do :D
 
swarleystinson said:
I just bought Metro 2033 (only $20 on steam!) and Crysis 2.
I'll download these, benchmark them, and let you guys know the results :D
Click to expand...
Sweet... just in case you don't know. Metro 2033 has an actual benchmark program that is installed in its application directory. (C > Program Files (x86) > Steam > steamapps > common > metro 2033 > metro2033benchmark.exe)

I'm particularly interested at seeing a benchmark at 1440p w/ DX11 on Very High with AAA, AF 4X, and DoF, Tessellation and Adv. PhysX disabled.
 
MythicFrost said:
Sweet... just in case you don't know. Metro 2033 has an actual benchmark program that is installed in its application directory. (C > Program Files (x86) > Steam > steamapps > common > metro 2033 > metro2033benchmark.exe)

I'm particularly interested at seeing a benchmark at 1440p w/ DX11 on Very High with AAA, AF 4X, and DoF, Tessellation and Adv. PhysX disabled.
Click to expand...

Thanks for the tip. I'll run it now. Of note, however. The game initially was running beautifully at 1920x1080p, 'high' settings DX11. It had a small black border on the sides when running but looked stunning. This was on standard clocks. I exited the program and upped the clocks to 800/1100 and when I restarted, the game was running completely fullscreen and was running like poo. In order to get avg FPSs in the 50s (min >30) I had to run 'normal' at 1650x1080! Tried restarting and the same results. Not sure what's up.
I'll run the benchmark, but if this is any indication, there's no way in hell i'll get anywhere near playable FPS on those settings.
 
As I figured.
With OC 800/1100:
Settings on 'High' at 1440p with AAA and PhysX disabled DX11 DoF and tesselation on...

Average Framerate: 15.00
Max. Framerate: 24.55
Min. Framerate: 3.11

I'll play around and see what I can get.
 
Mmm, interesting... In case I wasn't clear enough, I meant Adv. PhysX, Tess, and DoF all turned off, not just Adv. PhysX. The latter two are real FPS killers! :eek:
 
With 800/1100 OC...

Options: Resolution: 1920 x 1080; DirectX: DirectX 10; Quality: Medium; Antialiasing: AAA; Texture filtering: AF 4X; Advanced PhysX: Disabled; Tesselation: Not supported; DOF: Not supported

Average Framerate: 67.00
Max. Framerate: 133.56
Min. Framerate: 6.11


Options: Resolution: 1920 x 1080; DirectX: DirectX 10; Quality: High; Antialiasing: AAA; Texture filtering: AF 4X; Advanced PhysX: Disabled; Tesselation: Not supported; DOF: Not supported

Average Framerate: 48.00
Max. Framerate: 113.07
Min. Framerate: 10.71

Options: Resolution: 1920 x 1080; DirectX: DirectX 10; Quality: High; Antialiasing: AAA; Texture filtering: AF 16X; Advanced PhysX: Disabled; Tesselation: Not supported; DOF: Not supported

Average Framerate: 47.00
Max. Framerate: 92.71
Min. Framerate: 7.90

Options: Resolution: 2560 x 1440; DirectX: DirectX 10; Quality: High; Antialiasing: AAA; Texture filtering: AF 16X; Advanced PhysX: Disabled; Tesselation: Not supported; DOF: Not supported

Average Framerate: 32.00
Max. Framerate: 55.14
Min. Framerate: 5.24

...sorry bud. You're gonna need a PC gaming rig if you wanna play DX11 max settings.
 
MythicFrost said:
Mmm, interesting... In case I wasn't clear enough, I meant Adv. PhysX, Tess, and DoF all turned off, not just Adv. PhysX. The latter two are real FPS killers! :eek:
Click to expand...

Options: Resolution: 1920 x 1080; DirectX: DirectX 11; Quality: High; Antialiasing: AAA; Texture filtering: AF 16X; Advanced PhysX: Disabled; Tesselation: Enabled; DOF: Disabled

Average Framerate: 43.00
Max. Framerate: 86.90
Min. Framerate: 5.87

Can't turn tesselation off in DX11 (would sort of defeat the purpose!) But you're right, DoF is a helluva GPU killer.


I wish I could figure out what happened the 1st time I launched the game. It was like windowed mode but without seeing windows in the background. It ran really well!

2011-05-26 00:29:33 - Metro2033
Frames: 17192 - Time: 320s - Avg: 53.725 - Min: 29 - Max: 97
 
Those are pretty good. 32 FPS is playable, and even Medium settings aren't too bad on Metro 2033. I score 15 FPS with the same settings w/ my 4850M.

Is it possible to run a DX11 test but with Tessellation and DoF turned off, or are they forced on in DX11? From what I've learned, DX11 can actually improve frame rate, which is why I ask.

If possible, could you run the last test again (on post #39) but with Medium settings instead of High. And, after that, the same test but with 4x MSAA. If not though, no worries.

EDIT: Just saw your newer post, WOW... that's a big difference and pretty great with Tess on.

Cheers
 
You played this Metro 2033 on STEAM, right?

I wonder if we can get better result using Bootcamp windows? Well, someone out there, please share your experience :)

But I think it's okay, not all game, even modern one, made as heavy as Metro or Crysis. I cant help but wonder about Battlefield3 though, would it be more demanding than Crysis 2 or Metro? :(
 
iSayuSay said:
You played this Metro 2033 on STEAM, right?

I wonder if we can get better result using Bootcamp windows? Well, someone out there, please share your experience :)
Click to expand...
He is using Windows, Metro 2033 isn't available on Mac.
But I think it's okay, not all game, even modern one, made as heavy as Metro or Crysis. I cant help but wonder about Battlefield3 though, would it be more demanding than Crysis 2 or Metro? :(
Click to expand...
Being able to play on High at 1080p or Medium at 1440p is still pretty good though, and BF3 might be more demanding, but I doubt by much if that is the case.
 
MythicFrost said:
Those are pretty good. 32 FPS is playable, and even Medium settings aren't too bad on Metro 2033. I score 15 FPS with the same settings w/ my 4850M.

Is it possible to run a DX11 test but with Tessellation and DoF turned off, or are they forced on in DX11? From what I've learned, DX11 can actually improve frame rate, which is why I ask.

If possible, could you run the last test again (on post #39) but with Medium settings instead of High. And, after that, the same test but with 4x MSAA. If not though, no worries.

EDIT: Just saw your newer post, WOW... that's a big difference and pretty great with Tess on.

Cheers
Click to expand...

Is that good or do you want more?

Also, Just updated 1st post w/ Crysis 2 results. What I consider the settings w/ best graphics vs good FPS.
 
swarleystinson said:
Is that good or do you want more?

Also, Just updated 1st post w/ Crysis 2 results. What I consider the settings w/ best graphics vs good FPS.
Click to expand...
I checked out the first post, they're same nice results and really good improvements from the OC. I'm excited :D

If you could do one more test, I'd like to see one at 1440p w/ Tessellation, no DoF and no Adv. PhysX, Medium quality, AAA. And then the same again but with 4x MSAA. I'm hoping the game will be playable with 4x MSAA. And I'm interested to see how demanding 4x MSAA is over AAA.
 
swarleystinson said:
GREAT NEWS!!! (for those looking to overclock)...

Had an hour or 2 to noodle around today and gave my first attempt w/ an alternate program, Sapphire Trixx.

Default clock settings were 680 core/900 mem.
I was able to safely increase to 830 core/1130 mem with Tmax about 74C immediately after stopping benchmark.
Artifact/Benchmark testing was performed via Furmark v1.8 at native resolution w/ no AA.
I saw a jump in frames per second (FPS) from ~30 to 40 FPS.

I ran continuous testing for about 30 minutes max.
Fans were controlled via SMC fan control (though can be set w/ Trixx as well). Speeds were ran extra high as Furmark can cause really high GPU temps. I ran fans #s 1,2, & 3, respectively at 2200, 3000, and 2200 rpm.

Did not try and "real world" testing w/ games but will give y ou updates results on gaming improvements noted. Based on furmark and the numbers I'm expected a ~10-12 fps gain. Will post benchmarks for COD 4,6,7; deadspace2; the witcher; etc. Let me know if there's a game you really want me to test, I have a few.
I don't own crysis 2 but could check a demo.
Click to expand...

nice results
 
MythicFrost said:
Yeah, but the fans also run really slow. So notching them up compensates.
Click to expand...

Spot on. Fresh out of the box, at stock clock speeds, the top 27" iMac (w/ the 6970M) was hot to the touch after half-an-hour or so of The Witcher 2 in Windows.

Overclocking it to 800/1100 (a bit faster than desktop 6850 clocks) and setting the fan speeds to 2000/3500/2000 results in a somewhat louder system—though not uncomfortably so—and nets a 10% gain or so in frame rate, no crashes, and a much cooler machine to the touch. GPU temps, as reported by Sappire TRIXX & GPU-Z, don't budge over 70C after an hour or so of gaming.
 
Hey swarleystinson
Thanks alot for this.

I'm a first time OC:er and tried OC at stock voltage (i.e. 710/1000). However that ain't giving me any increase at all in the game I'm playing (IL-2 Sturmovik: Cliffs of Dover).
Noticed that the card is fairly cool when cranking up the fans so would like to try your OC.
But what voltage did you go for and could you briefly describe the process?

I know I should up the core clock first in increments, then the mem clock but dunno how to work the voltage.

Btw, what drivers are you on?


I tried ATT first but it locked the comp. Then MSI Afterburner but the new clocks wouldn't stick in game.
I use ATT for monitoring in game.
 
Last edited:
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back