Thank you MVP!
Do you think that PCIe v2 offers a noticeably increase in performance with a cMP 3.1 and a 750 ti or 970 versus PCIe v1?
As I have posted before, the 3,1 hits a glass ceiling with OpenGl and GPUs.
OpenCl and CUDA may not be as had hit by this.
With a 750 I doubt that the PCIE speed will have an impact.
You wouldn't be possibly interested in sharing those drivers or pointing to the post that told you how to do it, would you?
Hi, sure.
Here's the post I followed written by theBostonian
nVidia have not released the 10.9.5 version yet so no one can mod them.
FYI the actual modding is very easy, once nVidia releases their updated driver I'll try and do it myself
Actually, anyone comfortable with the terminal can do it:
Expand the .pkg
Code:
pkgutil --expand ~/Desktop/WebDriver-334.01.01f01.pkg ~/Desktop/WebDriver
use TextWrangler or other editor to edit var found_hardware = 0 to
Code:
var found_hardware = 1
Delete old .pkg then run this command:
Code:
pkgutil --flatten ~/Desktop/WebDriver ~/Desktop/WebDriver-334.01.01f01.pkg
[/B]
If need though, i'm happy to share the driver i've already patched. Can't guarantee it will work or won't harm your system though, although I imagine that's not likely.
Cheers
^^^^From what I've been reading, the only difference with this driver vs. the f01 drivers is that it fixed and removed the "Cuda Update Required" message that showed up on boot and when the Cuda CP was queried.
Yes, mine as well. Good to see that Nvidia is attending to business!Well, version f03 of the driver out today, and this one did remove the "Cuda Update Required" message required, at least on my first restart
http://www.nvidia.com/download/driverResults.aspx/79077/en-us
Lou
Thanks. I just checked & my card has 2xDVI & an HDMI socket. The point is that with the Nvidia web drivers both DVI sockets work just not both at the same time. With the default OS X driver they both work fine together. I don't have any performance issue that I think may be fixed by the Nvidia web drivers but I am guessing that they must be better/faster in some way else why would Nvidia make them available?^^^^At one time I was using a MVC flashed GTX 570, and it only had 2 working outputs - The lower DVI and the Display Port. I never used it unflashed. I still have it in the box. I put it up for sale in the Market Section and had no takers.
Lou
Yes, mine as well. Good to see that Nvidia is attending to business!
Edit - I spoke too soon. After the first reboot, the CUDA 'Update Required' text re-appeared. And, CUDA is still non-functional.
I have the same issue with my GTX 680. "Update required" and CUDA non-functional.
Is this something we're waiting on Nvidia to resolve with an update?? Or would you say this is something the "community" will have to find a solution for???