Hello...I was wondering if someone might be able to help me diagnose a problem. Today, suddenly, my entire screen was covered in what I assume are artifacts. I have encountered one or two before during stress testing in the past, but nothing like this. This was the entire screen. The colors were completely distorted and were a mix of green/pink lines/dots. My pc did not crash but restarting did not resolve the issue. I was simply browsing the internet at the time, nothing demanding, nor had I been doing anything prior. This was pretty much a cold-boot.
I re-seated the GPU and this seemed to resolve the issue. However, a few hours later during a session of Battlefield 3 the same problem occured. I tried uninstalling the drivers, ran driver sweeper in safe mode and re-installed. Still no fix. I tried removing the OC I had with MSI Afterburner, still no luck.
I tried re-seating but this did nothing to fix it this time. I eventually switched to another PCI-E slot and the problem was gone. I ran furmark for about 10 minutes, encounterered no problems and switched back to the original PCI-E slot. Artifacts were gone, ran furmark for another 10 minutes, things were peachy.
I've never seen anything like this before...I'm sure there is a problem with my hardware, but I have no idea what.
I haven't installed anything lately or done anything out of the ordinary. Computer was running fine yesterday. This is a new build that was completed in late October. I monitor my temps regularly, cores rarely go above 60 and running furmark my 580 tops off at 83 (I'm giving high estimates as well). Been running the 285.79 beta drivers for a couple of weeks now.
What can cause this kind of problem? What kind of tests should I be running? Thanks for any ideas and suggestions, this was just so sudden. Everything seems fine right now but there's no reason for it to be fine (I really haven't done anything), so I'm assuming whatever is wrong is not fixed.
I think you should download HWMonitor from CPUID and monitor your temperatures. It sounds like your cooling setup is flawed.
I could buy that if this had been happening under some sort of load. This was from a cold-boot, only thing open was google chrome. Aside from the one time that I was playing BF3, the other times it has been under normal use. If I could consistently replicate this problem by stress testing I would be considering temps, but this just seems to happen randomly. I've run multiple sessions of furmark with no success. I'll try running Prime95 later.
My case is a Silverstone Raven 03, and my idle cpu temps are below 30. I do monitor CPU temps and GPU temps through CoreTemp and MSI respectively.
Seems to be what I've been seeing online. I guess my big question is, would both cause random full-screen artifacts like that? Luckily my friend just built a similar pc so I can probably test using his components. I guess I'll try running my card in his rig to see what happens. Could this in any way be a motherboard issue? That's my big fear right now.
Thanks, that's a bit of a relief. I actually just scaled back my drivers to be safe. Odd that this could be the problem after weeks, but maybe some new conflict emerged. What bothers me the most is not being able to replicate this problem. If it is GPU related, would it more than likely be a problem associated with the VRAM? I hear furmark sucks at stressing that, so maybe that's why I haven't been able to replicate the problem using it.
I should mention that nvidia includes drivers that scale back the performance of the video cards when the Furmark or OCCT programs are detected running on the computer.
It would prevent your video card from being maxed out by Furmark. Furmark wouldn't be able to reach into the top 1/3 of the card's performance capabilities.
That is regardless of how good furmark is with VRAM or things like that.
Wow, thank you, I had no clue that was the case. While it's premature to celebrate, I've been running bf3 for about an hour and have yet to encounter any problems. I'm going to just assume that maybe it was the drivers, but I honestly have no idea. This could just be the early stages of a failing card and it could very well creep back. I sure hope not =(
TBH, though, that is the kinda thing you have to expect from beta drivers. They are still in a testing mode after all. Even the company that made them doesn't feel confident enough in them at this point to release them as is.
If they were completely stable it probably wouldn't be a beta driver.
Not that I am trying to guarantee the problem is fixed or anything, but I think scaling back the driver has a good chance of resolving the issue at least.