Random BSODs with random errors/bugchecks in random processes

Hello everyone,

I'm stuck with a very frustrating BSOD problem and from the look of it, it seems like some kind of harware failure. Thing is, I've replaced almost everything but the BSODs refuse to go.

The BSODs are completely random but playing a current generation game (Arkham Asylum, GRID) would guarantee a crash within an hour. Mass Effect, though, runs for many hours at a stretch sometimes...
Analyzing my dumps told me that the crashes were totally random. Page faults, the irq_not_less_than_equal one, the driver_irq_not_less_than_equal one, and the bugcheck codes were pretty random too! Even the faulting processes were anything from ntoskrnl to npfs to my ati driver... just about anything!

I like fixing my own problems, plus I'm impatient. My temperatures were within limits, so I assumed it was the RAM and got myself 2 new corsair 2GB chips just to be sure.
But the BSODs continued...

So it's not the RAM; the motherboard, Graphic card, and the PSU are brand new anyway. The only thing left was the processor, but they almost never go bad... or so I thought.

The real clue was when one of the dumps was found to be corrupt by the debugger! Reading up on that told me that it could be a processor issue. So I borrowed a Xeon 3040 from a friend to test.

Well guess what... The BSODs have almost gone. At least I can work in peace now though the Xeon is painfully slow. I say 'almost' because I DID get one BSOD since I put the Xeon about a week back. My friend tells me that I could have mismatched ram timings, and that my old processor might not be at fault at all!

@Route44: I need help. I want to get rid of the Xeon and get something faster. However, how can I make sure that it was a processor issue in the first place?! Can you help me check my ram timings or whatever else it might be? I still did get ONE BSOD even with the Xeon. What if I get another processor but the BSODs continue?!

I've attached the last few minidumps. The latest file is the one that happened with the Xeon. I'll be so so grateful if someone can figure annything out from these dumps. I have a feeling I've spent a LOT of money on an otherwise easily fixable problem... :blackeye:
 

Attachments

  • Minidumps.zip
    152.4 KB · Views: 1
Your dumps came up with errors 0x8E / 0xD1 / 0xA etc. each dump citing different / random process being responsible for instance in question.

Such errors do and can occur due to mismatch RAM timings or incorrect voltage settings. I'll advise you to check specs of your (i.e. timings + voltage) and see whether settings in BIOS are correctly matching these.

Secondly, sometimes even the settings are matching you may still need to fiddle around with voltage settings (incrementally) and see whether that stabilize the system.

Lastly, do stress test your graphic card with OCCT, because one of the dumps came up with 0xc5 as 1st argument, which usually happen due to acceleration issues of graphics. Regards
 
Thanks for the quick reply!

So basically what I have to check for is the ram timings and if that doesn't fix it then increase ram voltages till I get stable, right?

It'll take me a while to do these checks because my old processor is not with me and the Xeon replacement is not BSODing anyway.

But that brings a question; wouldn't I still be seeing BSODs if it was a ram timing issue? The Xeon does have a different FSB speed (1066Mhz as against 1333Mhz with the old processor). Could that be why I'm not seeing crashes now?
 
I download ooct like you said and ran it on the current Xeon anyway. I got an error on core#0 within 3 minutes itself. What's funny is that my FSB is being shown by ooct to be overclocked to 266.7Mhz instead of the Xeon's 266Mhz. That's an overclock of just 0.3%. Could that be a problem?

The graphic card test is currently running...
 
No 266.7 is something which I wouldn't worry about. Just ensure that voltage settings of Processor are according to specs.
 
Back