Memory Issues?

Looks like you've zeroed in on the problem.

Put one stick in at a time and run memtest on each to find which stick, or sticks, is bad. Then replace it (or them).

Yeah, I thought so too.

But after trying stick 4, which was what memtest suggested was the iffy one, that ran 4 passes without errors, so I stuck all the sticks back in, and tried to find an option in the bios to raise the timings, since the ram is sold as 5-5-5-18, but memtest is reporting it as running at 4-4-4-12. Couldn't find any options to change it though. Only ram option I could find was for the frequency, which can only be set at 750Mhz, twice the FSB, al the other options were larger multiples of the fsb, and much higher than the 800Mhz that the ram is rated at.

Anyway, given that the apparently iffy stick passed fine, I'm running memtest on all 4 sticks again, and its currently managed 3 passes without issue, further than it got last time.

So, rather than spending my evening running memtest, I've booted back into windows going to run a full memtest overnight again, should get enough passes to clarify everything.
Could memtest have made a false positive?

I don't know much about computers and I really have no clue what's wrong with your memory, but I just wanted to add to this thread something that may or may not be of much use to your situation. I play FFXI online and when I had my last computer, I used to get that same "the instruction at 0x000... referenced memory at 0x00.... The memory could not be read" message when I tried to run Guncam Video Capture software with FFXI. Parsing gives me similiar problems with memory sometimes. If you're running a Video Capture or Parsing program or something that interacts with WOW, that might have contributed to the problem. I never did get Guncam to work right after that with FFXI, so I use Fraps now. Mine also did not lead to BSOD at the time either. Hope that helps, but ignore if it doesn't :P

So, 6 1/2 hours of memtesting all 4 sticks together, 7 passes completed, and no errors showed up. Could this mean the 251 errors I got abover were false positives? Or is the problem just really obscure and hiding? I'm not sure :x

I've attached the two minidumps from the bluescreens. The one from the 4th was the MEMORY_MANAGEMENT one, and the 5th was the dxgmms1.sys not sure if there's any useful info in them to be gleaned or not, and I'm not sure what to test.
Think I'll load up WoW and see if I can make it crash again :x

And bellaneph, thanks for the reply. However, I'm not running any video capture programs, though I think I may have had a movie running on my second monitor at the times of the crashes. Not sure if its relevant or not. Thanks for the input either way ^_^

Each diagnostic test offers clues and brings you closer to the cause.

Multiple errors are not to be ignored. That is a real problem. Memtest has been known to pass bad RAM sticks, but you hardly ever hear of it giving failures without there being a problem of some sort.

But here is what may have happened: By moving the sticks around, plugging them in and out, shuffling them around, they are now making better contact in their slots. Poorly seated sticks, or just oxidation on the contacts can cause errata. So you may have gotten lucky and inadvertently fixed it already.

In addition, the sticks can be good but one or more of the motherboard sockets can be bad. Also, the voltages to the sticks can be wrong, miss-set, or fluctuating due to power supply issues. You can look at the voltage to the memory in your BIOS and compare it to the rated voltage for your sticks.

Running 2 or more programs concurrently can cause problems, particularly if one or more of the programs are poorly written and manage memory badly. You want to test to see if the problem occurs only when that condition exists. You can also check the game's website's forum to see if this is a common problem with that game.

You could also test the system for a while with all BIOS settings at their default speeds and/or frequencies. No overclocking or speed/timing changes. If you find that the system is stable at the motherboard defaults but failures occur as you increase the settings then that will help zero in on a cause.

Finally, you could try to stress test your system. You can use a program like Prime95 for this test: Free Software - GIMPS You can run the "Stress Test" for a few hours or overnight. This will not tell you what the problem is, but it is helpful to uncover any issues your system has with instability and cooling.

Hope that helps.

Goddamnit.

This time, the errors seem to be all over the shop? Rather than one paticular address like last time. Now, I'm confused by this, as the only thing I've done since last time was change the timings, from the 4-4-4-12 that it was running at, to the 5-5-5-18 that they are designed (Though thinkig about it, 5-5-5-15 would probably be better), but either way, the timings are alower, so the ram should be less stressed, and less prone to errors?

I'll try completely removing the overclocks and testing again today.

iviv said:

Hi,

I've had several crashes yesterday and today, and I'm a little stumped by them.

It started while I was playing WoW, bluescreen out of nowhere. No file specifically mentioned, but the error was MEMORY_MANAGEMENT, first part of the error code was 0x0000001a.
Shrugged it off, a one off thing's not too bad, loaded WoW back up, after 20 mins WoW crashed:
The instruction at 0x005595fc referenced memory at 0xffffffff. The memory could not be read.

Now, two crashes in quick succession, both talking about memory got me a little worried, downloaded memtest86+ and gave that a run. Completed one pass fine, wanted to play some more games, and the computer was stable the rest of the evening.

Now, today I was playing WoW and it crashed yet again, this time I screenshotted it:

Once again mentioning memory issue.

And finally, a couple of hours later, playing TF2 this time I had another bluescreen, this time it said the error was with dxgmms1.sys

Googling that particular file led me to these forums, which is why I'm posting here now. I had the idea to give memtest a proper long run while I was posting this (on my laptop) as that'll probably take a few hours to get a good number of passes in just to confirm its not a RAM issue, but of course that means I can't actually get hold of the minidumps windows made with the bluescreens... I'll post those once the memory's checked out XD

With all the memory errors getting mentiond, I thought that the issue would be with the ram, but so far it seems to pass memtest perfectly, so I'm a little stumped by what the problem could be. With the dxgmms1.sys bluescreen, that's made me consider that it could be a problem with the ram on the graphics card? I'm not sure if there's programs to test that ram like with memtest, but a few weeks ago I was running 3dmark Vantage and Heaven graphics benchmarks with no errors as well.

So... any ideas without the minidumps?

Edit:


Thread answered itself :x

Now what? Do I have to try the ram sticks 1 at a time to find the faulty one? Or is there a way from reading through all that junk to see which one is the dead stick?

Edit2: Right, I've removed all but the 4th stick, given the errors were happening at 4263Mb, but its run three passes of memtest without any errors now.

If you can get us the dmp files we can be of more use. If not since one of the DMP referenced dxgmms1.sys, I would remove all traces of the current video driver and then install





a fresh copu


Ken

zigzag3143 said:

If you can get us the dmp files we can be of more use. If not since one of the DMP referenced dxgmms1.sys, I would remove all traces of the current video driver and then install





a fresh copu


Ken

Hi, I attached the dmp files a couple of posts up.

Almost 3 hours into a memtest with everythng at stock speeds. And for reference, here are the memory clocks I used for the previous memtest, the one with 19 errors I just posted: