New ram memtest error

Toggle sidebar Toggle sidebar
C

camur.camur3

Jun 28, 2018
15
0
10
Hello. So I almost figured out why I get BSOD for a week now. I thought it was my GPU but this time it looks like it's the RAM modules.

I ran the memtest and I got over 100 errors. The RAMs are brand new. I did everything like uninstall/install drivers. Re install win 10 and everything. But the problem continues.

My rams are gskill tridentz rgb 2400mhz 16 GB, 8x 2.
Mobo brand new: Msi z270l gaming pro carbon ac. All settings are on default and auto, no OC. The ram is on 1.2v ,XMP is off and bios are up to date, what could be the problem ? Please help
 
Solution
InvalidError
If you have errors in memtest, one or more chips on your DIMM may be defective and the only fix for that is RMA the DIMMs for warranty replacements.
Sort by date Sort by votes
Start by choosing one of the XMP profiles for the memory modules, if they support it. It should correctly configure the modules voltage and timing. Rerun the tests after enabling XMP.

If you continue to get errors while using an XMP profile, you can either attempt to manually tune the memory or return the set for a replacement. You can power off the machine, remove one module, and re-run the memory test on a single module at a time if you want to know which module is defective.

The final test would be, after identifying a defective module, install the module into a different slot, preferably the slot the working module was in, unless both modules show errors. The point of this is to ensure it's the module and not the slot on the motherboard causing the errors.

If both modules return errors when run by themselves, I would suspect either a really bad configuration or possibly a defective memory controller on the CPU. Might be worth trying a different CPU at that point, if that's even an option.

Expect to return both modules, as dual channel kits are matched and need to be replaced as a set. There's no reason to do too much diagnostic work when you start getting memory errors in a memory test, especially when the RAM modules are new. The errors are almost always from a module being defective.
 
Upvote 0 Downvote
It's not unheard of for inexperienced users to accidentally bend pins in the CPU socket when installing CPUs. This leads to malfunctions from time to time, such as memory errors. Testing a module that is throwing errors in a different slot will help you determine if it's the module or the slot. If it's the slot causing the errors, it might be worth carefully removing the CPU and examining the socket for bent pins, or just using a pair of slots on the board that work without error.
 
Upvote 0 Downvote

When you have an important bent pin, the error rate will usually be so high that the computer won't even POST and if you do get into memtest86, errors will be scrolling by as fast as memtest86 is able to print them. OP on the other hand is managing to get into Windows and (un)install drivers. Re-seating components is still worth a try since it doesn't cost anything, but I wouldn't hold my breath.

A single hard bit error would generate hundreds of message over the course of a few seconds and let the computer run seemingly perfectly fine until an important piece of code or data lands on the bad bit and gets used. Since the computer is somewhat usable, I'm fairly confident the issue is weak/dead cell(s) on one of the DIMMs. I wouldn't recommend actually using the system in this state though as the bad RAM may cause the OS and software to corrupt data and program files.

In memtest86, bad cells can be confirmed by errors always landing on the same bits at specific addresses. A bad address line connection would cause massive multi-bits mismatches at random addresses due to the DRAM chip(s) being in the wrong place while bad data line connections would cause the affected bits to mismatch at random addresses. I'd be very surprised to see a computer manage to boot into an OS while the RAM is scrambling random bits and addresses. Even memtest86 likely crash repeatedly in those conditions.
 
Upvote 0 Downvote
When I try to play a game I play for maybe max 4-5 minutes before I get crash to desktop or BSOD. Also when I stress the computer. The temp of cpu is fine, it's very low.

When running memtest: I run it for 5 minutes I chose stop on error. First time I ran it I got 109 errors, second time I got 719 errors. I hope it's not the cpu that is bendt because I have checked it before.

Hopefully I will find out today what the problems are.
 
Upvote 0 Downvote
Hello, the topic got solved but it is not solved so i want to ask you here.

So I ran the memtest on both rams one by one and there was no errors. I ran them both for 20 mins, usually the errors occurred after 5 mins. Both tested on same slot on the left side.
And when I wanted to test the other slot I realized that the 24 pin connector is pushing the ram aside softly maybe 5° (not hardly).
The backside of the 24 pin is a bit thick and the ram modules are also thick. Psu is EVGA 600w non modular and drams are gskill trident z. Do you think this is causing the problem for the memtest to run unstable?
Thank you
 
Upvote 0 Downvote
If the modules are testing fine when there is no interference from the power cable, it sounds like you should do something about the power cable. They are usually stiff and unwieldy, but they can usually be managed enough that you can get it away from your memory module.

Moving the power connector far enough to not put any pressure on your memory modules and then testing the slot sounds like the next logical step.
 
Upvote 0 Downvote

The problem with "doing something about the power cable" is that the 24pin connector on the motherboard is right next to the DIMM slot and cannot be moved anywhere, which is why the connector is interfering with the wider RGB RAM heatsinks.

The simplest solution here would be to return the RGB DIMMs and swap them out for DIMMs with thin heat spreaders.
 
Upvote 1 Downvote
Personally, I would sooner replace the poorly designed motherboard than run on single channel memory.

You should be able to purchase RAM modules that are no wider than the motherboard slot they occupy. If your 24-pin power connection is too big to allow use of that slot with modules the same width or thinner than the module slots, your 24-pin power lead is the problem.

There are a few low-profile right angle 24-pin ATX adapters available, if you can wait for shipping from China.

You may also have enough flexibility with a 24-pin power extension, if the lead coming from your power supply is the problem.
 
Upvote 0 Downvote


 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom