I'm getting that old chestnut of a new build failure: no video output. I've tried every DVI-D on the machine (and with two 6970s plus onboard, that's quite a few). I've ensured both the 24 pin and the 8 pin are connected to the motherboard. I've tried starting it with non-critical devices disconnected. I've tried removing and reinserting the CMOS battery. I've tried two different monitors. The machine has power; I can get the disc drive to open and shut, and the diagnostic LED reads 00.
I dunno why they even list a code if it's description is "not used" but if the 0 you see is actually a D then take a look at the chip and see if there is damage or not seated properly...also take a look at the mobo socket. I'll search around for more stuff on this. check the psu and make sure your cable 8 pin is in all the way.....display to me though sounds like a gpu problem though...may be bad card or bad pcie lane...try moving it to another location. Also try and HDMI connection...i have heard that some mobo's wont display through DVI to initial BIOS until it has been updated...I though did not have that problem....
All four power cables are getting plugged in to the video cards, yes. The six-pins plus the little wonky two-pin things, and then the normal six-pins.
I juggled cards and slots, etcetera, but most recently I've been trying to use the DVI-D ports from the onboard. I'll try other cables and other ports, just in case, but I'm not hearing speaker beeps and the poster isn't changing. It really does feel like it just won't POST, rather than being a display problem... disc drive has also stopped working.
Eight-pin for CPU power seems to be in fine. In fact, I removed it and turned it on to see if the symptoms were the same, and they weren't--starting the machine without that CPU power causes the LEDs not to even turn on.
As for the CPU itself, well... see for yourself.
Here's the socket... as far as I can tell, nothing's out of place and no pins are bent.
Here's the top of the CPU. What you see here is the result of following the Tom's Hardware guide, i.e. put little dots of thermal paste around the middle evenly and the CPU cooler will spread it out. It definitely spread out, but it's pretty uneven. Could this be where I'm getting problems?
And here's the bottom of the CPU. Nothing seems wrong, no thermal paste.
Took it out, checked it, seated it again. I don't think it's seating improperly... the socket is keyed so that there's really only one way it can sit. Using the wire to lock the cover latch over the CPU itself seems like it's putting a LOT of pressure down on it, but I figured that was just the way it was supposed to work. Is there a chance it's doing damage? The only points where the latch is making contact with the CPU is over on the sides. It doesn't seem harmful, but who knows?
ADDENDUM: Just tried powering the motherboard and CPU with my old (lower wattage) PSU. Exact same symptoms. Bad CPU/motherboard?
The chance of it being a bad cpu are about 1/500. Chance of you having shorted out the cpu by applying too much thermal paste, about 1/5.
Not sure who your trying to blame by saying:
"What you see here is the result of following the Tom's Hardware guide, i.e. put little dots of thermal paste around the middle evenly and the CPU cooler will spread it out. It definitely spread out, but it's pretty uneven. Could this be where I'm getting problems?"
You apply the thermal paste by the directions on the package, not by some guide written by someone on TH. All thermal pastes are different material, different viscocity, different characteristics, therefore they all have different directions for application.
You can't use a Ford repair manual to re-build the engine on a Chevrolet.
Instructions provided with the paste were barely in English, and the diagram displayed the stuff pouring out into a perfect circle. It was just a little packet of the stuff that came packaged with the Xigmatek cooler. I used as much as was provided, which I assumed was the appropriate amount for the surface.
The guide, for reference, was this one. I'm not trying to "blame" anyone, I'm just saying whose instructions I worked from to see whether that was the source of my problem.
I wasn't aware shorting out a CPU with paste was even possible. Aren't they supposed to be non-conductive?
I suppose I'll be trying this again, then. Or having someone else do it.