Card: MSI R9 270x Gaming 2GB PSU: SuperFlower Leadex GOLD 750W Have spent most of Tuesday evening and yesterday afternoon right through to this morning trying to resolve some pretty serious GPU hard lock issues that have appeared out of nowhere that promptly escalated to full display artefacting the minute Windows boots up to the login screen and full lock-up 20-30 seconds after. Everything AMD cleaned off the machine and I can at least use the machine without the GPU crapping out, but all that's good for is browsing the net, everything else kind of requires proper display drivers to be installed as... you know... display drivers are kind of important. But otherwise problems so far have ranged from display corruption and machine hard locks, to Windows refusing to even boot until a system restore to pre-driver installation is done. Following things have been attempted already: * RAM, Hard Drive, PSU etc... all check out fine. * Partitioned off a chunk of my Hard Drive to do a fresh OS install and still immediately get the problem the moment anything other than the Windows default 'Standard VGA Adapter' drivers hit it. * Tried 10 different drivers including 2 year old drivers that the manufacturer still ships with the card. * Checked CPU and Memory clock speeds to make sure some update hadn't botched them up, and made sure they were kept to stock levels at their peak. * Checked GPU voltage and even tried under volting slightly in case something REALLY fecked up had happened with a driver update. * Updated vendor motherboard/Chipset drivers. * Updated vendor PCI adapter drivers. * And even did a CMOS reset in case the problem was on a BIOS level. And nothing improved things. Some observations: * GPU temperature from what brief seconds I can check before the problems kicks in seems fine. Around 60 degrees. * GPU Load is non-existent when the problem happens. * CPU and Memory clockspeeds are not outside of stock levels. * GPU voltage is stock (Or rather what software is reporting as stock voltage.) Lower voltage feels like it does buy me a extra 15-20 seconds or so before full death kicks in. Looking into people with similar cards discussing things they've noted their cards handle voltage 1.3-1.5 higher than what mines set at, so I'd say my reported stock voltage is definitely within safety margins. Only thing I can guess has happened is the memory on the card has worn out either something that's been on-going perhaps from the moment I bought the card back in February and it just suddenly concluded with spontaneous crap out, or something in the last week or so caused the voltage fed into the card to spike at one time or another and fried the memory very quickly. Any other ideas on things to try for those technical folk out there that may read this, I'm open to even the most obscure possibility to investigate that might disprove that the cards pretty much scrap now. I suppose I could whack on a Linux distro and see if for some crazy reason this is some random Windows-only problem that has magically appeared out of nowhere on what has been stable the rest of the time.. but that's really hopeful thinking.