There's only two things I can think of, and it's a combination of both. First, I've had problems with Gigabyte in the past. All sorts of strange issues, sometimes undefinable. Which is why I try to avoid this brand now all together. It seems Gigabyte boards sometimes really don't like certain hardware. I've never been able to find out why though... I try to use Asus now when I can, and maybe my next board will be Supermicro. Stability is more important to me than raw speed. The second thing that comes to mind is that maybe Windows automatically installed some update your motherboard doesn't agree with. Which could be why everything worked OK up to a few months ago and a clean install (with all the most recent updates no doubt) didn't fix it. I'm sorry I can't be more precise, but this is what my gut feeling is telling me.
Amusingly enough, I went TO Gigabyte from Asus for exactly the same reason
I used to have a lot of Asus hardware, but about ten years back I went through three systems in a row that blew up with no cause I could determine, in one case literally, taking the hard drives, ram, processor, and graphics card with it. Then ate another processor when I tried a spare in the board to see if it was a processor or board that was at fault (before I'd found everything else that was bad or I wouldn't have risked it!) which led to the obvious conclusion that the motherboard was faulty. So that processor went into the new board it had come from in the first place...
Two dead motherboards, two CPUs, a couple of gigs of ram when that was both a lot and expensive, three hard drives and a graphics card all went in the rubbish. Not happy.
And it wasn't a power supply fault either, the PSU was working fine when tested. It got chucked as well on the basis of being guilty by association, even so. Asus sort of admitted, without admitting it in a way that would lead to a refund, that there were some problems with that model of motherboard. Went right off them at that point
As far as this problem goes, I may have found the fault, or more accurately, faults. I'm almost certain the original graphics card has some mental issues of its own, but the root cause seems to be the ram disk driver I've been using for three years now not being compatible with windows 7! So your theory of a microsoft update breaking something is most likely true.
I certainly am not stupid enough to allow auto-updates in the house, but I do occasionally manually install security patches and the like. My guess is that one of these has helpfully cocked something up that the ramdisk program relies on, in a very subtle manner. It then interferes with the graphics drivers as well, possibly everything all ends up sitting in the same part of memory under some conditions when booting or something like that.
The end result is that if the ramdisk driver is installed, the machine BSODs on boot about 75% of the time, and runs flawlessly the rest of the time. Uninstall it, it boots (so far) 100% of the time like it used to. Put it back, it breaks again. And so on.
It's the most repeatable problem I've managed to find but only time will tell if it's the true cause. The problem is obviously that there's no way to know if you've actually fixed the issue, only that you've fixed it up to the point of the last boot. Who knows what the next one will do?
RobS and I spent about five hours on skype discussing the problem, deciding that it had to be a kernel-mode driver, then finding a tool to list them all and uninstalling everything that we couldn't see was essential. We got eight programs in before it would boot reliably four times in a row, which was the start of the solution.
Fingers crossed, but it may now work. We'll see...
I certainly hope it is, though, I was going slowly nuts trying to figure it out!
Previous Topic
Index
