Bad memory?

Sphexi

Diamond Member
Feb 22, 2005
7,280
0
0
Well, I have a 512MBx2 DDR Gold OCZ kit in my comp, plus a 1GBx2 Gold OCZ kit I just added in the past month. I've got fairly good cooling, the MB temp is never over 102F, and the ambient temp in the case is rarely more than 5-10F higher than the room temp. Since I put in the new memory I've noticed a lot of errors with programs dying, giving a message about not being able to read an address in the memory.

I downloaded Memtest and ran it full tilt for about 8 hours, two different instances each about 1023MB or so, and over about 800% worth of tests got a total of 3 errors. It seemed very random though, so I went into my bios and dropped the memory speed from 800 to 747mhz, and started memtest back up. They're both at 500% right now, 0 errors so far.

My question is, does this seem more like a mobo issue, or memory? I've got an Abit P5N32-SLI board, which I've found is horrible for overclocking (as in it won't at all), so perhaps it doesn't like running the memory at it's clock speed? Or does having 4 memory DIMMs in there create a heat issue? Or is the new memory just crap?
 
 

Sphexi

Diamond Member
Feb 22, 2005
7,280
0
0
Originally posted by: jjj807
try using one stick at a time

I'm going to have to, but can't until my next days off on Tuesday or so. Until then it'd just be nice to have an idea, as I'm running into the 30 days mark to return the mem here.
 

ryderOCZ

Senior member
Feb 2, 2005
482
0
76
Sphexi,

You have DDR2 in that board right? Not DDR. That would also be an Asus board not Abit if you have the P5N32-SLI. This is not a Core2Duo machine, right?

Do you have the voltage set to 1.9 - 2.0V or just the stock 1.8V of the machine?

Running 4 sticks in the system places a higher load on the memory controller, plus you are running 2 different densities which can cause issues in itself.

Try the voltage increase, also try increasing the NB core voltage to 1.3 - 1.4V. Try running just the 2x1GB sticks to see if the issues continue.

Another thought, is this mostly when you exit Internet explorer that you get this error?
"The memory referenced at XDSAJHGSKIFDSHGS cannot be read" is that the message?
 

Sphexi

Diamond Member
Feb 22, 2005
7,280
0
0
Originally posted by: ryderOCZ
Sphexi,

You have DDR2 in that board right? Not DDR. That would also be an Asus board not Abit if you have the P5N32-SLI. This is not a Core2Duo machine, right?

Do you have the voltage set to 1.9 - 2.0V or just the stock 1.8V of the machine?

Running 4 sticks in the system places a higher load on the memory controller, plus you are running 2 different densities which can cause issues in itself.

Try the voltage increase, also try increasing the NB core voltage to 1.3 - 1.4V. Try running just the 2x1GB sticks to see if the issues continue.

Another thought, is this mostly when you exit Internet explorer that you get this error?
"The memory referenced at XDSAJHGSKIFDSHGS cannot be read" is that the message?

I apologize, it is DDR2, and it is a C2D board yes. The message you posted is the one I'm seeing, it's not a blue screen, just an error message from the application itself, here's the message straight from the event viewer:

Application popup: Explorer.EXE - Application Error : The instruction at "0x03ea0f60" referenced memory at "0x00000080". The memory could not be "read".

Like I said, it's completely random as to which applications it targets, and I haven't seen it once yet since I clocked the memory down. I'll reboot and get all of the settings/voltages/timings.

EDIT:

Rebooted and checked, the mem voltage was 1.8, upped it to 1.95, and the NB seemed to be on Auto, the lowest this board will go manually is 1.4 so I left it there. I put the memory speed back to 800, and the timings are all set on auto, so I'm not sure how to tell what those are right now. I ran memtest again, two instances at 1023MB each, over 400% each with 0 errors so it seems to have resolved the issue. Thanks for the input!