Video card or RAM causing BSOD?

A

Avengeance

Gawd
Joined
Apr 10, 2004
Messages
773
Ok so I have had my computer setup for a while and I am trying to figure out what the heck is causing my BSODs. My specs are:

Core 2 Duo 6550
Intel DG33BU Motherboard
2gb OCZ PC6400 memory
XFX 8800GT Alpha Dog Edition
Antec Earthwatts 550W

So I sent in the video card about a year ago for service and they sent me a new one. The old one would give me BSODs and artifacting. I have always had a problem (it seems) with artifacting with this setup. I tried doing ram tests and even running it with another stick but the BSODs are so random its hard to test. The ram tests said the ram was fine after running it for a few hours. I guess my best bet would be to buy new ram and test it for a while and see if its the ram. I have a feeling its the video card though, I just dont trust XFX that much. Not sure if they have got better.

BTW, I always use a anti static wrist strap and time I touch anything inside of my computer.
 
What is the BSOD message?
Check your event logs to see if anything happens just before the BSODs.

Test your CPU with Orthos set for large FFTs and test your ram with Memtest.
Report back your max CPU load temp using Realtemp.
 
Last edited:
The message is IRQ_EQUAL_OR_LESS

Where does it even store the logs? I have tried to check before but never found them.

I guess I will have to do that with the CPU. I hope its not the CPU though. How long should I run memtest for? Would starting it at 11pm and ending at like 4pm the next day work?
 
A few hours should be enough to establish if there is a general problem.
The CPU max temp is of significance, please tell what Realtemp reports.

Event logs are stored in
Control Panel/Administrative Tools/Event Viewer
You can tell when a crash happens because the next entries are from Windows booting again.
 
Ok so I did a little prime 95 stress test. Started it last night and got up this morning and my computer was at the login screen. So I guess I just figured out that it wasnt my video card. It game me a little log file called "prime.txt" not sure what it all means but here is what it says if its worth anything. By the way I was doing "in place large FFTs" stress test.

V24OptionsConverted=1
SendAllFactorData=1
WGUID_version=2
StressTester=1
UsePrimenet=0
MinTortureFFT=128
MaxTortureFFT=1024
TortureMem=8
TortureTime=15

[PrimeNet]
Debug=1
Click to expand...
 
I think the IRQ_EQUAL_OR_LESS message usually means bad RAM or possibly too little voltage, so it's probably not the CPU.

Have you checked in your BIOS that the RAM is running on the correct voltage and clock/timings?
 
+1 on the ram.

If it was a video card you would have a bluescreen with the driver name at the bottom starting with NV.
 
There's no IRQ_EQUAL_OR_LESS bug check code. I'll assume you mean IRQL_NOT_LESS_OR_EQUAL. If my assumption is correct, then this message does not include or exclude RAM from being the problem. It simply means that a device driver or the kernel itself accessed paged memory under a privilege level that was too high. Touching paged memory might cause a page fault, and a page fault can't be handled when the faulting code is already inside certain parts of the kernel or device driver stack and running at a high privilege level.

Almost always, IRQL_N_L_E is caused by buggy code in drivers. The code is simply touching memory that it shouldn't. In rare cases, corruption in memory can cause code to wander off to visit the wrong memory, but that's not quite as likely as a bug in the driver itself.

If you post minidumps from the problem, I can have a look at them for you. But if you're running MEMTEST or the Microsoft Memory test for any reasonable amount of time without errors, odds are that memory isn't your problem.

If you've always had problems with artifacts, the problem causing that is almost certainly the video card. You have to determine if the video card's flakiness is also causing the bug checks you're seeing.
 
Thanks mikeblas, make sense to me.

I have tested ram some but I am going to test it more. I have it running right now and I will post when I get home what came of it.

My video card had artifacting problems but I sent it to XFX and they replaced it with a new one. The new one hasnt had any problems but I still have BSODs. Last night though my computer restarted itself while running prime95 with the large FFTs test.

Also, how do I get the mini dump files? I am pretty sure I already set it to make mini dump files but I am not sure where to locate them. Once I find one I will post it up here for you to check it out.

Thanks for all of the help guys!
 
Avengeance said:
My video card had artifacting problems but I sent it to XFX and they replaced it with a new one. The new one hasnt had any problems but I still have BSODs.
Click to expand...
I see. That's an important clarificaiton, as your original post you said you've always had problems with artifacts.

Avengeance said:
Also, how do I get the mini dump files?
Click to expand...
I wrote this post a couple of summers ago, and it has instructions and lots of background information on troubleshooting that you might find helpful. I don't know why it isn't a sticky anymore, but you've got a link to it now.
 
You should also try running Prime95 or IntelBurnTest a few minutes while actually watching it, and looking closely at what temps your CPU is running at( using RealTemp or some such ).
 
I never checked temps but I just figured they were fine but I will set up a monitor and do prime 95 overnight and see what it gets up to.
 
Ok, got home and the memtest had restarted probably many times. I guess that means its a ram issue.
 
Put only one stick of RAM in at a time and run memtest. They're prob not both bad. Also check in BIOS to make sure your mobo is detecting it properly. I know I almost always have to manually set up voltages and latency.
 
Yea im going to do the one stick at a time test. But as far as voltages go I dont think that I can change the voltage on this board. This board is good for just keeping everything stock lol.
 
why don't you take out your gpu and run prime95 with the igp. see what happens.
 
tricky0502 said:
and how old is your psu?

when my old psu was on its way out i was having these kinds of problems.
Click to expand...

I have my eye on this also but need to see the results of other tests which are slooow coming.
 
Not until we know the state of your system and if there are problems that can be resolved, unless you have another PSU.
If you have another known good PSU that is powerful enough, try that and see if it helps, but without another PSU the process is slower to troubleshoot.

If you dont do the tests we need, progress will be slower.
 
Havent had time to take out the 8800gt yet I think I am going to do that tonight and test it overnight with prime95. But I was wondering is it a chance its still ram if my comp crashes with prime95 large ffts and also memtest fails or does that seem more like a voltage issue or motherboard issue?
 
Just test with a single stick of ram and take it from there. If the tests pass then its the ram, if it fails with one stick try it with the other stick.. if it passes, well then you know where the problem is..
 
Tested with prime95 without the 8800gt in and it rebooted itself while I was gone so it failed. So I guess I can rule out the power supply and video card. Well I guess its more unlikely that its the power supply and for sure not the video card causing a problem. So now I need to just go buy some ram and see if it fixes it. Good thing Fry's has a good return policy.
 
Ok so I re ran memtest without the video card in with each stick of ram one at a time. Both have done over 7 hours of testing with no errors at all.

So now I am wondering what is really wrong with the comp.
 
I've had RAM pass memtest and still be the problem.. within the last month in fact. My Mushkin RAM I originally had passed memtest 3x over but my system was still locking up and BSOD crashing. Replacing RAM fixed it.

Have you looked in BIOS to see what it's setting your RAM to? It could just not be set properly or it could still be the RAM.
 
Ram is set to 5-5-5-12 T2 1.8v as far as I remember. There are the photos on page 1.

I looked up the settings for the ram, what it needs and its spot on. I checked the voltages under full load in HW monitor and they are all right. I dont know what could be wrong.

I think I am going to try prime 95 with different ram in there just to test it.
 
Last edited:
OCZ memory did this for me too. I had the same exact problems as you. I ran memtest for hours but the system was just unstable even at their recommended volts + timings. I put in a different type of memory and all my issues went away. I would return that memory for a different brand myself.
 
TheResidentEvil said:
OCZ memory did this for me too. I had the same exact problems as you. I ran memtest for hours but the system was just unstable even at their recommended volts + timings. I put in a different type of memory and all my issues went away. I would return that memory for a different brand myself.
Click to expand...

No need to go different brand, you might just got stuck with a bum set, OCZ is a very good company. Also, I had 8800 ultras and the ram on those went, BSOD all day for me. Changed em out and now golden.
 
Might not be related, but a couple of your temps are rather high, specifically the memory controller hub( MCH ) at 65C and the TMPIN5 at 68C. If that's at idle, which appears to be the case, it's possible those things could be overheating when under load.
 
What are their failing points? The case is well ventilated and I have all the cables cleaned up so it shouldnt get too hot in there. Maybe the mobo is flawed?

That would make sense with the problems I am having. Seems like when I have both sticks in it fails memtest but one stick it does fine. Maybe running dual channel heats it up faster?
 
maybe you should go into the bios and load default setting. maybe somehow some of your volts got changed?
 
Cant change volt settings on this mobo as far as I know. I already reset it earlier because I had disabled the IGP so I had to reset the cmos so it would come up when I was testing the comp without my vid card in.
 
Ok satus update: I ran prime 95 on the "blend" test, which tests memory mostly. With both sticks in I got a BSOD with the error "an attempt was made to execute non-executable memory" and so I tried one stick at a time. One stick at a time seems to work fine. So why would both sticks at once cause a problem but one at a time is fine?
 
Maybe you havent set the voltage as per the rams spec?
 
Ram is set to spec. I took some pics and posted them on the first page. I am running 1.8v for the ram and thats all it asks for. I cant change any voltage settings or FSB settings or any of that. This board is very limited.
 
Have you tried using different slots for the memory? One of the slots could be bad, switch them over to the other two slots (assuming you've ruled out bad ram).
 
Actually I havent tried that yet lol I probably should. I always put them in slot 1 and 3 but when I test one stick I only put one in slot 1 so slot 3 could be bad.
 
You must log in or register to reply here.
Back
Top