I have no idea what Blast QE is. Certainly not a distributed computing project. There are no BOINC applications that leverage a Cross-fired or SLI'd gpu configuration that I know of or heard of.
I wanted to update the community on the results of my tweaking. I accomplished the goal of reducing the temps in my hotter running GPU card with a combination of two things.
I reduced the peak core voltage of the hotter running GPU from the default of 1.15V down to 1.1V using Afterburner. If I go any lower, the card starts to become unstable and crashes. This helped reduce the GPU temp across all w/u types.
The other thing was, I robbed some CPU computing power from the 1.22 G/R w/u when they run. I had the utilization factors set for 0.5 CPU and 1.0 GPU. Using those factors, when running 2 G/R w/u, 1 of my 4 CPU cores was set aside to assist with the GPU computing. I changed the CPU factor to 0.3. Now the system doesn't set aside a full core. It does take a little longer to complete a G/R GPU w/u, but my temps are more reasonable. I'll accept that trade-off.
Just looked at my completed w/u. I seem to have a bunch of Validate Error w/u isolated to the computer in question (12771426). I bumped up the under-voltage on the one GPU card to see if that solves the issue. Keeping my fingers crossed.
EDIT!
OOPS again!
Looks like the issue was with #2 GPU card. I may have over-clocked it a little too much. Dialed it back a little. Lets see what happens.
I have no idea what Blast QE
)
I have no idea what Blast QE is. Certainly not a distributed computing project. There are no BOINC applications that leverage a Cross-fired or SLI'd gpu configuration that I know of or heard of.
I wanted to update the
)
I wanted to update the community on the results of my tweaking. I accomplished the goal of reducing the temps in my hotter running GPU card with a combination of two things.
I reduced the peak core voltage of the hotter running GPU from the default of 1.15V down to 1.1V using Afterburner. If I go any lower, the card starts to become unstable and crashes. This helped reduce the GPU temp across all w/u types.
The other thing was, I robbed some CPU computing power from the 1.22 G/R w/u when they run. I had the utilization factors set for 0.5 CPU and 1.0 GPU. Using those factors, when running 2 G/R w/u, 1 of my 4 CPU cores was set aside to assist with the GPU computing. I changed the CPU factor to 0.3. Now the system doesn't set aside a full core. It does take a little longer to complete a G/R GPU w/u, but my temps are more reasonable. I'll accept that trade-off.
OBTW, I still have crossfire disabled.
Thank everyone for their help.
OOPS! Sorry wingmen!Just
)
OOPS! Sorry wingmen!
Just looked at my completed w/u. I seem to have a bunch of Validate Error w/u isolated to the computer in question (12771426). I bumped up the under-voltage on the one GPU card to see if that solves the issue. Keeping my fingers crossed.
EDIT!
OOPS again!
Looks like the issue was with #2 GPU card. I may have over-clocked it a little too much. Dialed it back a little. Lets see what happens.
Solved!!! Corrupted GPU driver update!