I'm checking out all the driver vers to find out when this problem originaly occurred.
I'm currently using 319.92, reason I wanna update it is because it's about a year old now.
Anyways it looks like the latest ver I can get my hand =s on without the prob might be 320.49.
I have not actually checked it though.
320.86 I know is affected, and it looks as if 320.78 is affected as well but I haven't checked.
I'm just going by what the readme's say this sec for those untested drivers.
I don't think it needs the edid forced because it can still read it from the monitor.
I'm gonna try this program called nvidia pixel clock patcher, it might take me a day to get around to trying it.
If that doesn't work, I'm try my hand at hex editing the driver and see if I can't figure out what changed.
I just get ver 320.49 downloaded and I'm gonna try that.
Um and in the driver readme's the issue is laid out as:
Multi-Monitor Support on GRID
Multi-monitor support on GRID boards K1
and K2 requires the following VBIOS
versions:
•
GRID K1:
80.07.AF.00.00
or later
•
GRID K2:
80:04:BA:00.00
or later
I read back when that people were having trouble getting past single link speeds in there virtual machines and this was the fix for it.
So I assume the fix and problem is one in the same.
Tesla k10 might be a solution, I haven't checked in a while which needs which resistor config but whatever is the easiest
.
These smd's, I've never seen them so small before, the ones on my card are sooooo tiny.
And thanks for the replies anyways
.