WDDM 1.3 issue resolved Hey ManuelG, I've just found the problem, the card is now reporting as WDDM 1.3 in dxdiag. The issue was caused by my onboard VGA, an intel HD 3000 (a WDDM 1.2 device). A very important note is that I've never used the HD 3000, and my display was directly connected to the 560 Ti. The HD 3000 does not appear in dxdiag as a display device either, it's just an inactive display adaptor so far as I can tell (it's only mention is it's drivers loading, full dxdiag.txt here - http://pastebin.com/DnHaL9Qr). I happened to be in the device manager rolling my 560 Ti driver back to 326.01 when I noticed the Intel HD 3000 in the display devices section, directly above the 560 Ti. Since I've never used it and wasn't even aware the device was available to windows, I decided I'd reboot and disable it in the bios just remove it as a factor in the WDDM problem. Upon reboot and explicitly disabling the HD 3000 in my bios, the 560 Ti reported as WDDM 1.3. I figure I should report this behaviour because: a) Intel HD 3000 is incredibly common, it's the sandy bridge onboard VGA. b) My motherboard BIOS default is to set the device to AUTO, which even with the primary display device set to PEG (PCI Express Graphics slot), still results in the HD 3000 becoming initialised (but inactive) in windows. The result is this is going to affect a lot of users with onboard sandy bridge VGA (ivy bridge HD 4000 has a WDDM 1.3 driver), whose motherboard bios defaults set the Onboard VGA to AUTO, even when their NVidia card is set as the primary video device. My particular motherboard vendor is gigabyte, for instance, and this issue is probably going to affect all sandy bridge users of their motherboards who are also using BIOS defaults with their NVidia cards. I had to explicitly disable the onboard VGA in my bios, such that windows never sees it, in order for the 560 Ti to be presented as WDDM 1.3. Disabling the HD 3000 only in window's device manager will result in the 560 Ti reporting as WDDM 1.2, so it has to be bios disabled. If I had to guess, windows is probably initialising the HD 3000 first, and as the WDDM 1.2 device that it is, then deactivating it since it has no displays attached, and then initialising any further display devices as WDDM 1.2 devices regardless of their actual WDDM capabilities. As another guess, this is probably a Microsoft issue rather than a NVidia one, and should probably be forward to them. Anyway thanks again for the fast responses and the confirmation that Fermi devices should be reporting as WDDM 1.3, ManuelG. After all that's what led me to persist and find the cause of my WDDM problem.