Wierd instability - odd memtest results?

Associate
Joined
16 Jan 2003
Posts
1,236
Location
by my pc. :D
Hi,

My machine has started to behave oddly - crashing unexpectedly when browsing the net, when not doing much etc. Crashed with a few different blue screen error messages too.

First thing I did was run memtest86 4.1 (off bootable usb stick) to make sure the memory was ok - and it started detecting errors. Odd thing was it said errors occured at 17272 MB - which is bizarre since I only have 16GB installed? Lower limit of errors was 5500 MB.

Dropped memory speed from 1600 to 1333 (basically disabled XMP) and reran tests - and it went through a couple of passes without issue.

Just been playing skyrim for a while and it's quit to desktop with no error messages or anything.

Reran memtest again - and it's showing errors again, this time at 17268.1 MB.?? Lower limit was 7424.3 MB.

I'll try swapping out the memory if I can find some spare - but any idea why I would get such a weird figure, and why the memory issue is supposedly intermittent?

-----------

Ran memtest86+ v4.20 (offshoot of original app) in an AMD pc with one of the two memory sticks - and it failed at 8298.3MB. On an 8 GB stick... something screwy on the addressing side of the memory module maybe?

Other module has tested fine in main rig.

Not having a good couple of days - must be something in the air (literally? like static? ;) ). RMA'd a pair of corsair modules from the amd pc as one was showing multiple errors. Issues today on my main rig with geil dragon ram. And I have a second set of geil ram which won't run at it's rated 16000 when the pair are used on their own.

Sky + HD box has started having issues with recording giving a "technical error 7"..

.. and I tried fitting an accelero extreme 7970 but even after following the instructions carefully the glue hasn't stuck when been left all day. :(

Yup. Must be something in the air. At least I fixed somebody else's laptop to balance things out a bit. :P
 
Last edited:
RMA'd the geil memory and received a replacement pair yesterday. Checked them over and they seem a little odd - on each one there is a chip that stands out from the rest - darker, and a different date/batch code on it. Either just mixed bins or possibly reworked memory?

Ran memtest86 and it tested fine for the first two passes but errors were reported on the 3rd pass. Heat related? Will try them in another pc tonight.

My corsair modules also came back from taiwan (replaced with a new pair as well) and worked fine when I tested them on arrival. Tried using those last night and one of the 4GB modules is completely dead and isn't recognised at all. :(

Not having much luck with memory lately......
 
I though I was having memory problems a few weeks back. I was going to pull the trigger on some new memory( and I almost wish I did looking at the the prices now). I was swapping the memory around on my x58 mobo testing each stick in different sockets. running it a lot slower than it's rated speed.

The results where seemingly random, I couln't pin the fault down.
as it Turns out.....

My north bridge heat sink was loose. Fixed that now it's as solid as a rock at normal speeds. are you fortunate enough to have another machine to test your memory on. and swap other componantes with?

Now the samsung quad channel kit's are pretty cheep here. might finally get some ddr3 that runs at it's rated speeds at 1.5v :)
 
Last edited:
Back
Top Bottom