Maximum PC

Recently updated my HTPC box with Windows 8. The ASRock A780GMH Motherboard's onboard video wouldn’t play nice with the new operating system so I purchased and installed an ASUS GT520 video card. I completely uninstalled the old ATI Radeon HD 3200 drivers and used Guru 3D Driver Sweeper for good measure. Installed the new card and the most recent drivers downloaded from the GeForce website (v310.70). New card worked flawlessly...for about 45 minutes. Then starting displaying flickering green and pink squares randomly all over the screen.

Checked the GPU temperature with Speedfan--it is hovering at about 62C, well under the max allowed 102 degrees.

Checked the power requirements for the card. Minimum required wattage is 300 according to GeForce's website: http://www.geforce.com/hardware/desktop ... ifications. I'm packing a 300W SeaSonic SS-300ES, so technically should be ok--if only just barely.

Here’s the rest of the hardware setup:
AMD Phenom II X3 2.5GHz
Kingston HyperX 8GB (4 x 2GB)
Western Digital Green 500GB SATA 3.0Gb/s HD
Mushkin 120GB SATA III SSD
LG Black Blu-ray/HD DVD-ROM
Dell S2309 Monitor (DVI) + Samsung TV (HDMI)

Any ideas on what might be causing my problem? Is it probably the PSU? Or is there another culprit I’m overlooking? Other troubleshooting steps I can take to identify the source of the problem?
Thanks in advance for your input.

Artifacting does sound like overheating, but I'd get a second opinion from HWMonitor.

I gave the system a 2hr "cooling off" period and on reboot the video signal is clean. This also makes it sound like a temperature problem. I'm dubious only because of the speedfan reading and because the ambient temperature in that room is about 63F. It just doesn't seem likely that it would overheat in those conditions does it?

I'll use HWMonitor to check temps if it happens again and report back.

Thanks.

I noticed that your PSU has a wattage rating of 300W, which seems as though it might be somewhat anemic, given the hardware that you are using in this machine. While your videocard is fanless and requires no additional power beyond the 75W that the PCI Express slot provides, I don't really think your videocard is overheating. And, since you mention that the onboard video on the motherboard did not work well with your new OS, I would venture to say that the motherboard might be defective.

Plugging all that hardware into http://extreme.outervision.com/PSUEngine says the power supply is more than fine for the job.

Update:

Testing with HWMonitor confirmed the temperature readings I was getting with Speedfan, so those were accurate.
I ran FurMark to stress test the GPU. Temps immediately climbed and peaked out at just over 100C and within seven minutes the artifacting/video noise began.
I shut down, cooled off, and rebooted. Then streamed Netflix full screen while monitoring temps. GPU temps never exceeded 85C but artifacting still occurs within 30 minutes.

Tomorrow I will borrow a 430W PSU from a work computer and test again. If artifacting persists, I will conclude that the GPU is junk and RMA. If artifacting disappears I guess I'll have to upgrade my system's PSU.

Thanks, Northwood for the link. That's a cool online tool that I wasn't aware of.

[EDIT] Pulled out the power supply and discovered it actually a 400W Corsair (much better than the 300W SeaSonic I thought I'd installed). So I don't think the PSU is the source of my problem.

After I put everything back together, I ran FURMark again. This time it ran 15 minutes, reaching 102C without any problems. Next I left it streaming Netflix and an hour later artifacts were back. Temperature was 85C. I'm out of ideas. Guess I will RMA the sucker.