Windows experience index fails, screen goes black and not responding.

Hello OVC forum,

I have a problem that I fixed before but it is back again..
I recently upgraded my GPU ( Sapphire Radeon 7870 ), before that my WEI was working, now I ran the new WEI and my screen goes black and I have to pull the power off because it isn't responding.

I tried deleting the WinSat files, result: screen blank again.
Tried it with CMD WinSat, exactly the same, screen black.
It always stops at the end when it reaches CPU assesment. Almost to the end, it really frustrates me.
Anyone has a solution?

Specs ( if necessary ):
AMD 8120 @ 4.0 GHz
8GB Corsair Vengeance
Sapphire 7870 Radeon
320GB HDD
Win7 Ultimate

Thanks in advance!

Does it happen with your CPU at stock speeds?

EarthDog said:

Does it happen with your CPU at stock speeds?

Click to expand...

Gonna try it now!

Already fixed, thanks man!
I feel so stupid now...

LOL... can you post what happened just in case others run across this problem?

FearDubstep said:

Already fixed, thanks man!
I feel so stupid now...

Click to expand...

Can you post what did you do to fix the problem?

I have upgraded my GPU to 7870 and facing the same problem.

This person has not logged on since the problem was fixed... I do not think you will get an answer from him.

EarthDog said:

This person has not logged on since the problem was fixed... I do not think you will get an answer from him.

Click to expand...

Hmm.
So, I am now hijacking this thread, as the problem is related to what my problem is and googling for such problem will land you here.


I have upgraded my GPU to MSI 7870 Ghz edition(2GB000).
My system specs are:
core2duo E4500
Abit IP35-E
Corsair Dominator DDR2 2*2GB
Seasonic S12 II 620W

When I try to stress test my system be it OCCT stress test/ Windows experince index/ GPUZ rendering test or play any game, my display goes blank, fan, HDD remain powered up, and I have to restart my system. I tested for voltage using multimeter, however multimeter shows constant output.

I have noticed that whenever GPU load increases, there is small spike to 0V on GPUs 12V voltage. So, most probably PSU is the culprit.
Here is screenshot:


^We can clearly see that, as load increases there is small spike in voltage.

PSU is a quality unit. Unless it is going bad, that is not the problem. Plenty of power to push that PC, even overclocked.

Are you overclocked? Perhaps it is not stable?

EarthDog said:

PSU is a quality unit. Unless it is going bad, that is not the problem. Plenty of power to push that PC, even overclocked.

Are you overclocked? Perhaps it is not stable?

Click to expand...

Nothing is overclocked. OCCT CPU stress test goes smoothly.

Please host your pictures internally following the instructions listed above...

The problem has been solved.

It was issue with tempratures!
After all testing and running around, If found out that it was high temprature that were causing the problems.

Here is the problem:

As I was running on bench mode, I used a cable to support the card(black braided cable that goes over the GPU cover on the right-top), so that the card can sit straight in the PCIE slot. The cable pulled the card from one side, which in turn made the heatsink bend and loose contact with GPU core. Now, loosing contact with GPU core is sureshot way of increasing temprature.

After I got rid of the cable, things are working fine. OCCT stress stess is running smoothing, I was able to play 30mins of Crysis2.

This is such a dumb thing!
n I struggled for two days doing all kind of tests and what not.