LinX error, can be caused by RAM ?

Grooveriding

Diamond Member
Dec 25, 2008
9,147
1,329
126
Hi folks,

Having an issue with random BSODs on my system in sig recently. I tore it down to add an EK MB waterblock to my loop and after putting it back together have been getting some BSODS. APC_INDEX_MISMATCH , IRQ_NOT_EQUAL , PAGE ERROR IN UNPAGED AREA are the usual text strings, with codes 05, 09, 24, 50 depending. My system is overclocked, but it is using the same settings that were rock solid prior to this starting.

Note, I also lined my case internals with this stuff: http://store.carbonfibergear.com/3m-di-noc-carbon-fiber-vinyl I have it lining the mobo tray as well. So I do have some suspicion there may be a grounding issue as my headphones are now giving me some feedback on mouse movement and hard drive activity which they didn't before.

I also have a suspicion that perhaps the mounting of the motherboard water block could be causing an issue ?

But, away from that - I have been failing linx with setting memory to 'all' and having all 16GB of memory installed (4x4GB) But, when I run with 1 stick, 2 stick or 3 sticks it passes linx just fine. I have isolated it down to adding one stick in particular that causes failure errors in linx and the BSODs to continue. Thing is that I passed 200% of hcimemtest with 12 instances running as well as memtest86+ overnight ? o_O

I realize there is a good chance it is just a bad stick of memory not showing in those tests, but the errors didn't occur until after I tore down and reassembled my case. Against that though, the errors only occur due to having this one stick installed. I've also tried swapping around different DIMM slots and still it is this one stick regardless of slot.

Just wondering if anyone thinks it could be something I am missing or is it just the obvious? - that it is this sneaky stick of RAM that avoids throwing off errors anywhere but in linx and random BSODs.

Maybe I just somehow borked this one stick in handling it ? I guess that is possible. Hoping it is just the RAM, will give me an excuse to pick up some of those 30nm samsung DIMMS that overclock like crazy :D
 

Idontcare

Elite Member
Oct 10, 1999
21,110
64
91
That's how I found out that I had a bad stick.

The memory passed memtest86+ for 24hrs so I figured it was good, but I kept getting LINX errors even with everything at stock.

Then I found out about HCI memtest, so I gave it a shot. Bang, memory error detected.

Replaced the stick (after figuring out which it was), confirmed it was good in HCI, and then no more issues with LINX.

Also, 200% in HCI is not much, you are supposed to shoot for 1000%.
 

AdamK47

Lifer
Oct 9, 1999
15,739
3,581
136
5GHz is a heck of an overclock for SB-E. I wouldn't be surprised if it was clocked just a bit too high.
 

Concillian

Diamond Member
May 26, 2004
3,751
8
81
Back when you OCed with FSB, I'd use LinX to approximate the limits of FSB, RAM and CPU separately, so yes, it will expose issues with any subsystem.
 

Rvenger

Elite Member <br> Super Moderator <br> Video Cards
Apr 6, 2004
6,283
5
81
IMO memtest86 is not as good as it used to be. I just use prime to detect the errors anymore.
 

Grooveriding

Diamond Member
Dec 25, 2008
9,147
1,329
126
Thank you for the input everyone. I found my problem, apparently I had never been running my RAM per Gskill's spec for my mobo. It worked fine before, but I guess maybe it flaked out.

https://docs.google.com/document/pub?id=17VKzOB5CL5EtIelw5tGsj2CqjBk3O0OU5VPe6CUoFso

This is how they said to run it and it eliminated my problems. I am guessing 3 of the sticks were okay running as I had them and the fourth was flaky. After making that change the system is stable again :rolleyes:
 

darkjairo777

Junior Member
Mar 19, 2013
1
0
0
Thank you for the input everyone. I found my problem, apparently I had never been running my RAM per Gskill's spec for my mobo. It worked fine before, but I guess maybe it flaked out.

https://docs.google.com/document/pub?id=17VKzOB5CL5EtIelw5tGsj2CqjBk3O0OU5VPe6CUoFso

This is how they said to run it and it eliminated my problems. I am guessing 3 of the sticks were okay running as I had them and the fourth was flaky. After making that change the system is stable again :rolleyes:


At the end of that stick of ram was bad or it was just a matter of configuration?

And were GSkill or corsair ram you have now.