BSOD/Memory errors, finding the solution?

Vidmaster23 · Apr 21, 2005, 11:15 PM

Well I am a hardcore gamer and play games a ton as you might imagine, mainly World of Warcraft.

Starting a couple months ago, I would ocationally get Memory errors that would crash the WoW client to desktop. Slowly until now, it has gotten more frequent, to the point of BSOD's with the STOP message. These errors are very strange for me. One day they will only happen once or twice. The next day they will happen 20 times (like once every 5-10 minutes, making games pretty much unplayable).

First thing I did was check all the fans, which were operating properly. The second thing that came to mind was a faulty driver. I hadn't formated in awhile so I gave it a shot. That didn't work. Next I installed different video card drivers, didnt work. I made sure I also had all the latest drivers for every component in my PC. Everything was good. I then suspected the RAM, so I ran MEMTEST and the Microsoft ram tester, both of those programs reported no errors.

I then decided to look at my CPU heat. Now here is where my first question comes in. While using Asus Probe 2 to check my CPU, I noticed my Athlon 51-FX was running WoW at 64 Degrees C. That sounded really hot to me, so I went and got a better fan (A Zalman fan). Put that baby on, I'm still running at 64 Degrees C. Is this Temp for an Athlon 51-FX normal for running CPU/graphics card dependant games? What should the temp really be? My idle temp is 47 Degrees C btw. And just to let you all know, my case is organized and clear. I have one fan in the front to bring air in, and one in the back to push it out.

My only other guess as to what's causing this problem is my graphics card, which is a Radeon 9700 Pro (had this baby forever). However, I'm leaning towards my CPU, cus that thing just seems to be running too hot to me, and no matter what I do it seems to stay that hot.

Does anything have any other things I could try to narrow down the problem? Thanks.

Vidmaster23 · Apr 21, 2005, 11:48 PM

Ran Prime95, this is what I came up with:

[Fri Apr 22 00:40:34 2005]
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.

FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.

First time I ran it, got that error. Second time I ran it, got that error again, same point, same time. The CPU also went up to 67 Degrees C during the tests. I did the Torture test In-place Large FFTS

FearTheTerp · Apr 22, 2005, 12:59 AM

It is a bit hot, pretty damn close to that 70 degree mark which in most computers would shut down automatically so if it didn't, then im guessing it is a RAM problem. BTW did you put the thermal paste right on your heatsink?

I would either loosen the timings on your RAM or if you have 2 sticks, take one out at a time and see if your games work with one and if you get the BSOD with another. You could also borrow a stick of RAM from a friend to see if the problem stops.

Tipstaff · Apr 22, 2005, 01:52 AM

Quote:

Originally Posted by FearTheTerp

It is a bit hot, pretty damn close to that 70 degree mark which in most computers would shut down automatically so if it didn't, then im guessing it is a RAM problem. BTW did you put the thermal paste right on your heatsink?

I would either loosen the timings on your RAM or if you have 2 sticks, take one out at a time and see if your games work with one and if you get the BSOD with another. You could also borrow a stick of RAM from a friend to see if the problem stops.

Agreed. That error usually pops up with shody ram, but if you are overclocking this error can crop up too. Now, even if you are not overclocking yourself, the ram might be running automatically above it's designed specs (Geil, at one time, was selling 333mhz DDR with an adjusted PID that made it run @ 400mhz.. according to them it 'ran stable' inhouse, so they figured it was fine to sell.. which was a load of crap.. and misleading.. plus it failed to run stable 1/2 the time).

As was mentioned I would try another stick of ram.

-Tip

mkk · Apr 22, 2005, 07:23 AM

64C in stressful conditions is nothing to worry about in my opinion. Sure it may not be ideal but we are still far off from harms way and the CPU won't get less stable from it either. Try what Tipstaff suggested, and I'd like to add that if you have tweaked any RAM settings in the BIOS then try returning them to standard values (SPD/AUTO).

Vidmaster23 · Apr 22, 2005, 03:56 PM

It seems as though whenever I just have one stick of RAM in the computer, it is very stable, and doesn't crash at all. I tried both sticks of ram by themselves and ran into no problems after playing WoW for a long period of time.

When I put both sticks back in and played WoW/Half-Life 2, I got a Memory error and a BSOD STOP error again.

What can be causing this? I have had 0 problems with dual memory until now. All of a sudden it's causing problems. As far as I know, both sticks are completely identical. Neither of the sticks has been overclocked, both have been at stock settings.

>GSXR<mrbusa · Apr 22, 2005, 04:17 PM

Quote:

Originally Posted by Vidmaster23

It seems as though whenever I just have one stick of RAM in the computer, it is very stable, and doesn't crash at all. I tried both sticks of ram by themselves and ran into no problems after playing WoW for a long period of time.

When I put both sticks back in and played WoW/Half-Life 2, I got a Memory error and a BSOD STOP error again.

What can be causing this? I have had 0 problems with dual memory until now. All of a sudden it's causing problems. As far as I know, both sticks are completely identical. Neither of the sticks has been overclocked, both have been at stock settings.

are you running your ram with spd enabled in your bios??if so then disable spd and set all your settings manually.make sure your ddr voltage is set correctly and monitor your psu and make sure there is not alot of fluctuation in your voltage output on any rail.

Firstperson · Apr 22, 2005, 04:59 PM

I had a similar problem with a amd 3200+, the cpu was bad, L2 cache problem, rma'd with AMD ...problem solved.

If there is anyway to test the cpu in another computer, try that, if the problem follows, its the cpu.

Firstperson

Vidmaster23 · Apr 24, 2005, 04:22 AM

Thanks guys for all the help. I got it fixed a couple days ago. I put both DIMMS in, went to the BIOS, and manually set all the latency and stuff in the BIOS instead of it being auto detected. There was one setting that was incorrect (at least acording to the Mushkin site) so I set that to the correct variable. So far I've gone 2 days without crashes.

Thanks again.

>GSXR<mrbusa · Apr 24, 2005, 11:53 AM

Quote:

Originally Posted by Vidmaster23

Thanks guys for all the help. I got it fixed a couple days ago. I put both DIMMS in, went to the BIOS, and manually set all the latency and stuff in the BIOS instead of it being auto detected. There was one setting that was incorrect (at least acording to the Mushkin site) so I set that to the correct variable. So far I've gone 2 days without crashes.

Thanks again.

np