TwoCables

But that thread said that it only happens with the NF200 chip. That is, 750i and up apparently. I did not see the nForce 600 series mentioned. But if it was, then I want to be corrected as soon as possible.
Nope. There's a reason why this is in the Intel - General forum. I mean, my GTX 260 isn't overclocked.
I have good news: I finally began the process of re-stabilizing my overclock. I began last night by seeing if it could pass 100 linpack runs using LinX. It passed all 100 runs using 3400 MB of memory (and it took 7 hours and 27 minutes). I tried to use more, but Windows 7 kept complaining about being low on memory as soon as I started the test. So, I kept lowering it by 10 MB at a time, and for some reason Windows stopped complaining when I got down to exactly 3400. Although, it still complained like an hour into the test. hehe But the overclock passed all 100 runs.

So, now that it has passed that test, I plan to keep going by using the 64-bit version of Prime95 to see if my overclock can do the following:

• 12 hours of the Small FFTs test.
• 12 hours of the Blend test.
• 12 hours of the Large, in-place FFTs test.

If it can do all of that, then I will finish it off by seeing if my memory can handle 10 runs of Memtest86+, as well as whatever might be the equivalent to that with the Windows Memory Diagnostic tool.

If it can do all of this, then I will let time tell me whether or not it's stable. If I still have the same problems (which are admittedly hard to describe since they're so infrequent), then I'll have to either make the overclock more stable, or consider a new motherboard since I would definitely like one that supports PCI-E 2.0.

So, let the stress testing games begin, or something! hehe
Unfortunately, my 9800 GTX+ is already sold to a friend who desperately needed it.
What's scary is I already have the Southbridge voltage bumped up to 1.675V. The lowest possible selection in the BIOS is 1.50V, and the highest is 1.750V. It used to be at 1.50V (Auto) back when I still had the 9800 GTX+, and it was also like that before I got my G.SKILL Falcon SSD. It was back when I was still only using my VelociRaptors (one for the OS, one for storage).

But then one day I bought Mirror's Edge (shortly after installing the official Release Candidate of Windows 7) and I kept experiencing the same crash at nearly the same exact point in the game almost every single time. So, after giving it some hard thought, I decided to increase that southbridge from 1.50V to 1.55V. The problem happened a little less often, so I bumped it up to 1.60V. Then the problem happened much less often. And actually, I was also experiencing a weird Firefox crash in Windows 7, and increasing the Southbridge voltage to 1.60 made this happen less often as well. So, I bumped it to 1.625V, and the problem seemed to go away completely.

But then I got my G.SKILL Falcon, and now that same crash in Mirror's Edge was happening at a much earlier point in the game. Oh man, it was like in the first level as opposed to about 3/4 of the way through the game. I was also experiencing other funky issues, so I increased the Southbridge voltage to 1.650V. That helped, but not completely, so I increased it to 1.675V which basically eliminated the problems once again.

And then the GTX 260 enters the picture. I'm considering increasing the southbridge voltage again, but I'm scared to actually use 1.700V. That's just a big-looking voltage. lol But the best part about all this is even at 1.675V, the southbridge remains cool at 44°C. I used to see 43°C at 1.50V.

What's even scarier to me is my Northbridge voltage is now at 1.50V, and the highest I can select in the BIOS is 1.55V. It used to be at 1.45V, but bumping it up to 1.50V seems to have stabilized everything (at least for now). Fortunately, it's running cool too at 53°C with the fan at 75% in the BIOS. And like with my Southbridge, 53°C is pretty much what I have always seen.

So, I dunno: but as long as it's stable and not burning up, I don't care what it takes - just as long as I don't have to spend any money on new cooling solutions.

__________________

Clockadile Dundee

How bizarre. Were you guys talking about a virus earlier?

Any BSOD codes or info that may be of use?

The fact that this comes back after an SSD... does that point to the southbridge, now - but not before!!?!?

LOL at bizarre instabilities.

r31ncarnat3d

This definitely sounds like a driver crash; did you install the right drivers?

I had this problem one time when I installed x86 drivers instead of x64. The card would run fine normally, but would always crash when I'd get into intensive gaming or even DVD playback.

__________________

TwoCables

Unfortunately, there is no virus involved. But fortunately, I think my system is 100% stable right now. The problems I mentioned before are gone so far. However, I still have a four tests left (Large, in-place FFTs, Memtest86+, HCi's MemTest, and Windows Memory Diagnostic).

But yeah, when I installed the SSD, increasing the southbridge voltage most certainly gave my system back it's stability.
Nah, I'm most definitely using the correct drivers. I'm too anal for that. I mean, I wish it were that easy hehe. But like, when I go to choose the driver on NVIDIA's website, I check the selections like 3 times before I begin downloading. And then as soon as the download starts, I check the filename to make sure it is clear that it's the right one. For example, I see it like this (well, kinda): "190.62_desktop_win7_winvista_64bit_english_whql.exe".

But I recently and unwillingly bumped the northbridge voltage to 1.50V from the previous voltage of 1.45V. Now it has just one more bump to go before I can't go any higher. I mean, 1.55V is the highest possible selection in the BIOS. Fortunately, the NB temp is still good. But more to the point: it feels like my system is much more stable now.

__________________

Clockadile Dundee

Well, let me know if I can do anything.

TwoCables

Oh, then give me some time and I'll think of something.

__________________