Hi all, My config is - Core 2 Duo E4500 2.2Ghz running at stock Gigabyte G41M-ES2L Corsair DDR2 2x2GB 800 Ram MSI 8800gt factory OC , Nvidia Driver 341.92 Corsair CX430w PSU 2 mechanical HDDs So I fired up Unreal Tournament 3 today after a long time on my old system and was trying out a few settings to make the game run smoothly. There's a setting called "Framerate smoothing' which basically uncapped the max framerate from 60 to whatever my gfx card could push out. When I did that I noticed that although the framerate at even the menu bumped up from 60 to 250, some ripples started showing on my monitor, running from bottom to top. Thinking it was a problem with my monitor I ignored it and just kept on fiddling around with the game. After some time, lo and behold!, my system artifacted and hung!! I restarted it and it was still showing me artifacts, from the BIOS all the way up to Windows. Thankfully, after reinstalling the driver the artifacts cleared up and it got back to normal. Thinking it was a software problem, I fired up UT3 again but with 'framerate smoothing' off, essentially capping it down to 62fps. I tried it for a few minutes and noticed that the ripples were showing up again. And, after a few minutes it artifacted! I restarted and tried Portal 2 and within a few minutes it happened again. So now i'm quite worried as to what could have happened and whether the damage is permanent. 1) Is my gfx card damaged that it can't handle high load games? 2) Is my PSU damaged? Its under warranty so I can get it replaced provided it is damaged. 3) Is some other part damaged? 4) The gfx card is 8 years old and has never had its thermal paste replaced. Its also been a few months since its been cleaned. I did check the GPU-Z log file and under load the temperatures didnt go above 70C .Should I try replacing the paste and cleaning the card up? Have never replaced the GPU paste so dont want to risk if its not needed. Thanks! Update - It artifacted and crashed while writing this post :/
Sounds like you have some dry joints on the card. If you can't afford to upgrade it, take the fan, heatsink etc off and stick it in the oven for 11mins at 220deg. Make sure the die is facing up. Once baked let it cool down for an hour then reapply new paste and put back together. This will be a temp solution that may last. As it's still working to a point this is your best chance of saving it
This is a last resort! The cheapest route is to dismantle the card and give it a good clean. That'd certainly involve replacing the thermal paste.
This is called 'tearing', and will ALWAYS occur when your rendered framerate is above the display refresh rate. When you are running at abnormally high framerates (north of 200 FPS) GPUs have very weird power behaviour. This often exhibits as 'coil whine' as the rapidly spiking load causes power circuitry to rapidly expand and contract from the rapidly cycling current. While a well-build GPU should not be permanently damaged by this, it may have been the last straw for an older GPU like the 8800gt. Turn framerate capping back on (all you're doing by turning it off is rendering extra frames you will either never see, or will only see parts of through tearing, and wasting GPU power) and see if the crashes continue.
Well I think I did hear more noise from the case, not sure if it were the cpu fans or the coil. I'll open it and clean it up and see if that helps.
Can it be a Psu issue worth doing a rma? A few months ago the computer was restarting abruptly but after a few months it just stopped doing that.