Problem with RAM slot #2

[TechPhoenix]

Hey guys,

I've been browsing around Tom's Hardware to find a solution for my RAM problem and I've got to the stage of using memtest. I've ran memtest on each stick in slot #1 and they ran successfully. As soon as I tested one of them in slot #2, I was flooded with errors from memtest. Right now the count is at 28928 errors, insane eh?

Now I probably figured that the problem is with the motherboard. I should mention that I ran 4GB of RAM in slot #1 for 2 years now. After about 1/2 days after installing another 4GB (of the same memory type) in slot #2 I've been on the receiving end of endless BSODs. I can run my computer fine with the 4GB in slot 1 but I would love to have the 8GB since they've helped a lot for the first few days.

Is there any solution that I can try myself? I have already contacted MSI about it but I'm still waiting for a response.

Quick update: My older memory stick seems to pass memtest without flaws. The newer one is the one that seems to report errors on slot #2. Neither of the memory sticks report errors on slot #1.

Here's the system specs:
MOBO: MSI H61M-P31-G3 Motherboard
GPU: XFX Radeon 7777 Black Edition
CPU: Intel Pentium G860
MEM: 2x Corsair CMV4GX3M1A1333C9 4GB
PSU: Corsair CX430
HDD: Seagate ST500DM002

Here's a photo of memtest:
Here

 

[house70]

Not quite. You can't pinpoint where the problem is, and even if you could, in case of a hardware-related issue (bad soldering, broken connectors, etc) you won't be able to fix it anyways. Avoid using that slot while waiting for an answer from MSI. They might take the high road and replace the mobo, or they might try to blame you for the failure (say, you pushed a RAM stick too hard into the slot and broke some connection).
Hardware-related problems are rare, but uncorrectable. I had a CPU that worked fine on 2 cores, but BSOD'd on four cores. Go figure; took me a while to pinpoint the problem, but replacing it solved it.

 

[TechPhoenix]

Figured as much...

Seems like my older RAM stick works correctly in slot #2 after a quick test.

Another test with both sticks installed (as shown below) seems to show no errors so far.
Slot 0: NEW stick
Slot 1: OLD stick

My older setup was:
Slot 0: OLD stick
Slot 1: NEW stick

The problem is I've tried this setup even before using memtest but I still encountered problems after day or two. It seems like switching between them alleviates the problem for a while. Any ideas?

 

[house70]

It could be that running the system for a while causes heat (obviously), which in turn expands some badly soldered joint and causes the problem.
Try to reproduce this by running a CPU-intensive task (like Prime95) immediately after switching the RAM sticks, when it should be OK. If errors thrown a lot sooner than anticipated, it's likely because the heat produced did it's job.

Another quick thought: make sure the mobo's firmware and BIOS are updated as to support that amount of RAM. I know it should not be an issue with any relatively new mobos, but it's better to be sure.

 

[TechPhoenix]

Not quite. You can't pinpoint where the problem is, and even if you could, in case of a hardware-related issue (bad soldering, broken connectors, etc) you won't be able to fix it anyways. Avoid using that slot while waiting for an answer from MSI. They might take the high road and replace the mobo, or they might try to blame you for the failure (say, you pushed a RAM stick too hard into the slot and broke some connection).
Hardware-related problems are rare, but uncorrectable. I had a CPU that worked fine on 2 cores, but BSOD'd on four cores. Go figure; took me a while to pinpoint the problem, but replacing it solved it.

It could be that running the system for a while causes heat (obviously), which in turn expands some badly soldered joint and causes the problem.
Try to reproduce this by running a CPU-intensive task (like Prime95) immediately after switching the RAM sticks, when it should be OK. If errors thrown a lot sooner than anticipated, it's likely because the heat produced did it's job.

Another quick thought: make sure the mobo's firmware and BIOS are updated as to support that amount of RAM. I know it should not be an issue with any relatively new mobos, but it's better to be sure.

I have never updated BIOS firmware and I sort of fear doing it since there is a potential that something could go wrong and I'm left without a BIOS. I might try it if it's really recommended (I'll have to check the firmware changelogs).

Funny thing is I didn't even have to use Prime95, a simple restart and running memtest again has brought the errors right back. I'm getting significantly less errors with 2 sticks but there still are around 464 at Test #8 right now. Heat sounds plausible but I'm not even sure, the temperature couldn't possibly rise that much just from memtest.

 

[TechPhoenix]

Seems like I misinterpreted the problem this whole time. The NEW RAM stick is the cause of the problem. I ran it on SLOT #1 and I got thousands of errors from it. It seems that sometimes it comes up with error and sometimes it doesn't. All things indicated that the problem came from SLOT #2 but it seems that the actual RAM itself is not working properly. I should have ran some more tests before posting but either way thank you very much for your advice. While I'm still not 100% certain, the fact that the older RAM slot can run more than 1/2 PASSES without errors on either slot seems to prove that the newer stick is the cause of all problems here.

Ah well, hopefully I can get a quick replacement for Amazon. Thanks again for the help!

 

[house70]

I was thinking about the RAM stick, but some details given were leading me in the wrong direction. Glad you sorted it out, enjoy!