Jump to content
LaptopVideo2Go Forums
zipper

Benchmark results for NVIDIA GeForce GTX 680M

Recommended Posts

Robbo the Second

Thanks for the reply Zipper.  When I had my M1530 I was able to mod the inf files myself too, but I'm a bit confused with my new system & the newer drivers when it comes to modding the inf files myself.  When you mod the inf file, do you use nvwin.inf, or one of the other inf files?  I believe each of the different inf files are specific to the manufacturer?  Does it matter which of the inf files we choose to modify?

Share this post


Link to post
Share on other sites
zipper

Preferably the MSI inf [NVMI(n).inf] but another will do - best to alter the lines for GTX 680M of another make. Edit the subsys or leave it out from the lines.

Share this post


Link to post
Share on other sites
Robbo the Second

You don't happen to know which inf is specific for Dell or Alienware do you?

 

EDIT:  It's OK, I did some googling, and worked out that it's the nvdmn.inf file that I need to modify.  I edited the 680M lines from the M17xR4, and turned them into 670MX.  Working great.

Edited by Robbo the Second

Share this post


Link to post
Share on other sites
zipper

Beta 331.40 tried with the original inf. No problems except GPU-Z still freezes. Quite a good one, worth trying.

Share this post


Link to post
Share on other sites
tribaljet

What's up with GPU-Z freezing on so many drivers? Being beta is no excuse for that behavior.

 

Also, does 331.40 have CUDA?

Edited by tribaljet

Share this post


Link to post
Share on other sites
zipper

GPU-Z just doesn't like this particular combo, 0.7.2 is OK. And 0.7.3 is OK on my older lappy. Must check that CUDA - I'd probably mentioned it if missing but not 100% sure. (checking..) GPU-Z 0.7.2 recognizes CUDA.

Share this post


Link to post
Share on other sites
tribaljet

I have to say that 327.23 is proving to be a very decent upgrade over 310.90, with increased performance nearly across the board. Just a shame that Nvidia decided to lock Mozilla-based browsers' plugin-container to the IGP on Optimus systems. However, it seems that results can vary from system to system, even on same generation GPUs, given that while I did get improved performance on my GT 555M, I've seen people with GT 525M report lower performance.

 

Could you give 327.23 a spin, just to compare performance with both 327.24 as well as 331.40?

Edited by tribaljet

Share this post


Link to post
Share on other sites
zipper

I did - almost identical, hint of slower than 327.24 but I've lost the results into a black hole (?)

Share this post


Link to post
Share on other sites
zipper

331.58 notebook WHQL tried. A pretty good performer, like the preceeding WHQLs;  worth trying.

Share this post


Link to post
Share on other sites
zipper

At last some decent overclock figures, clockdowns do happen still but not so early; 830/2460 instead of default 771/1800 give 6 to 11% increase depending on the benchmark used, most on Fire Strike. Updated to Win 8.1, don't know if it has any effect.

Share this post


Link to post
Share on other sites
tribaljet

I forgot to ask, did you get to try the 327.54 developer driver? Seems to be the latest driver available.

 

Also, why exactly isn't Nvidia shipping proper Windows 8.1 drivers but rather release Windows 8 drivers that happen to be compatible with Windows 8.1?

Share this post


Link to post
Share on other sites
zipper

331.58 WHQL is officially:

Operating System: Windows 7 64-bit, Windows 8.1 64-bit, Windows 8 64-bit

 

327.54 developer driver not yet, perhaps later.

Share this post


Link to post
Share on other sites
tribaljet

Regardless of public statements, it's a WDDM 1.2 driver, hence Windows 8 driver, not Windows 8.1 which has native WDDM 1.3.

 

EDIT: For some reason I've seen different Fermi GPUs having WDDM 1.2 and WDDM 1.3 with the same drivers, no idea what's going on there.

 

EDIT2: Actually, it seems all systems I've used report WDDM 1.2, whereas there were some reports on the web about 331.58 supposedly reporting WDDM 1.3, which I find hard to believe given that the driver base is still geared towards Windows 8, with marginal changes that allow for stable yet not actually optimized usage on Windows 8.1.

Edited by tribaljet

Share this post


Link to post
Share on other sites
zipper

Tried the MSI version of 326.80. To my surprise some benchmarks were clearly better than with Nvidia OEM 326.80 installed with MSI inf; 1.5 to 2.5 % is noticeably better.

Share this post


Link to post
Share on other sites
zipper

331.65 notebook WHQL now - about the same as 331.58 or a little faster on some benchmarks. No problems.

Share this post


Link to post
Share on other sites
tribaljet

You do realize you're exceptionally fast at benching drivers, right? Good job :amen:

Edited by tribaljet

Share this post


Link to post
Share on other sites
zipper

But it took several hours after the news...

Share this post


Link to post
Share on other sites
tribaljet

No buts! You saw, you installed, you benched. And it wasn't that long since drivers were released.

Share this post


Link to post
Share on other sites
tribaljet

On a different note, there are reports that 331.65 seems to have broken V-Sync? Can you confirm that?

Share this post


Link to post
Share on other sites
zipper

Haven't used it - must have a look.

Share this post


Link to post
Share on other sites
zipper

331.82 notebook WHQL tried - almost identical to the predecessor.

Share this post


Link to post
Share on other sites
tribaljet

Did it by any chance have working V-Sync?

 

Also, I should add that 327.41 are the most stable Windows 8.1 drivers for Optimus systems so far.

Share this post


Link to post
Share on other sites
zipper

Seems to work for me - but no CUDA re GPU-Z 0.7.4.

Share this post


Link to post
Share on other sites
zipper

331.93 notebook beta tried. About identical to the preceeding WHQL  - but 3DMark website doesn't like this at all. Mostly won't validate.

Share this post


Link to post
Share on other sites
zipper

332.21 notebook WHQL tried. No problems, good but not exceptional results except 3DM11 is lower than expected. And that CUDA bug is fixed at last in GPU-Z as well as Optimus detection.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

×
×
  • Create New...