Memtest
-
- Posts: 6926
- Joined: Thu Feb 10, 2000 8:00 am
A related question... I ran memtest on my machine after I got it built, but the results seemed inconclusive to me - or more precisely I probably didn't know exactly how to read them.
I'm not sure how many iterations it wound up running through before I stopped, but it was going for probably a couple hours.
So it listed a few problems, but I wasn't ever sure exactly what they meant. Is there some sort of key or guide that helps you read the results in a way that you could use to pinpoint problems?
Or is it just a diagnostic in the sense of "If there are any errors, then you have bad memory or memory slots?"
I'm not sure how many iterations it wound up running through before I stopped, but it was going for probably a couple hours.
So it listed a few problems, but I wasn't ever sure exactly what they meant. Is there some sort of key or guide that helps you read the results in a way that you could use to pinpoint problems?
Or is it just a diagnostic in the sense of "If there are any errors, then you have bad memory or memory slots?"
Mmmmkay...... I just ran memtest again with some very poor looking results.
Keep in mind the last time I ran this was right after I built my computer. I got a lot of errors, but at the time I was running dual-channel with 2 x 1Ghz sticks, and was still trying to find ram timings that would allow my system to run stable.
So now, I've got only a single 1Ghz stick in, running at default timings for the mobo. No dual-channel, no agressive timings.
I got 111 errors, and that is only running one single pass of all the tests.
When changing the error reporting mode to find bad ram patterns, this is what it shows:
BADRAM = 0x1000e790, 0x1000e790
Both of those are exactly the same. It started running a second pass of the tests before I could stop it, so that second one may be from the second pass. Either way, both of them being the same makes it seem like a bad block of memory on my stick, does it not?
Do you guys know anything about the test, who can tell me if my guesses are correct?
I'm about to try running this again with the stick in a different slot, and then maybe try running it with my other stick instead of this one.
I knew I should have gone with my tried-and-true Corsair RAM instead of OCZ.
Keep in mind the last time I ran this was right after I built my computer. I got a lot of errors, but at the time I was running dual-channel with 2 x 1Ghz sticks, and was still trying to find ram timings that would allow my system to run stable.
So now, I've got only a single 1Ghz stick in, running at default timings for the mobo. No dual-channel, no agressive timings.
I got 111 errors, and that is only running one single pass of all the tests.
When changing the error reporting mode to find bad ram patterns, this is what it shows:
BADRAM = 0x1000e790, 0x1000e790
Both of those are exactly the same. It started running a second pass of the tests before I could stop it, so that second one may be from the second pass. Either way, both of them being the same makes it seem like a bad block of memory on my stick, does it not?
Do you guys know anything about the test, who can tell me if my guesses are correct?
I'm about to try running this again with the stick in a different slot, and then maybe try running it with my other stick instead of this one.
I knew I should have gone with my tried-and-true Corsair RAM instead of OCZ.

Well. I put my stick in a different slot and ran the test again, with pretty much the same results as before, so it can't be a single slot that's bad.
Then, I swapped sticks and ran the test on my other 1gb stick. I got TONS more errors than the other stick - I'm talking 3500 errors without even letting the 3rd test finish.
I put it in a different slot as well, with the same results.
It sounds to me like OCZ has given me some shitty RAM. Keep in mind that these were two matched sticks that were supposed to have been tested in dual-channel mode and running at 2-3-2-5 timings (which is what I'm running at now by default).
I'm calling the shop I bought it from tomorrow to see if I can get my money back and just get Corsair.
Then, I swapped sticks and ran the test on my other 1gb stick. I got TONS more errors than the other stick - I'm talking 3500 errors without even letting the 3rd test finish.
I put it in a different slot as well, with the same results.
It sounds to me like OCZ has given me some shitty RAM. Keep in mind that these were two matched sticks that were supposed to have been tested in dual-channel mode and running at 2-3-2-5 timings (which is what I'm running at now by default).
I'm calling the shop I bought it from tomorrow to see if I can get my money back and just get Corsair.

Got this off the Memtest86 site:

Sounds exactly like what happened to me (see my bad HDD thread).Frequently memory vendors question if Memtest86 supports their particular memory type or a chipset. Memtest86 is designed to work with all memory types and all chipsets. Only support for ECC requires knowledge of the chipset.
All valid memory errors should be corrected. It is possible that a particular error will never show up in normal operation. However, operating with marginal memory is risky and can result in data loss and even disk corruption. Even if there is no overt indication of problems you cannot assume that your system is unaffected. Sometimes intermittent errors can cause problems that do not show up for a long time. You can be sure that Murphy will get you if you know about a memory error and ignore it.

-
- Posts: 6926
- Joined: Thu Feb 10, 2000 8:00 am
-
- Posts: 6926
- Joined: Thu Feb 10, 2000 8:00 am
-
- Posts: 4022
- Joined: Sat Mar 12, 2005 6:24 pm
Yea, that's kinda what I figured from looking at the program and the names of the tests. It seems to have the most errors during the test that writes a random number sequence.
I guess it could MAYBE be a problem with the mobo or processor, but after doing the stick/slot swaps that I did, I'm fairly certain the memory has something wrong with it.
I'm hoping OCZ support can give me some good news, but I'm not sure if I'll ever be able to recover the data from my HDD.
I guess it could MAYBE be a problem with the mobo or processor, but after doing the stick/slot swaps that I did, I'm fairly certain the memory has something wrong with it.
I'm hoping OCZ support can give me some good news, but I'm not sure if I'll ever be able to recover the data from my HDD.

I posted the full saga on the OCZ support boards:
http://www.bleedinedge.com/forum/showthread.php?t=18977
http://www.bleedinedge.com/forum/showthread.php?t=18977
There is also an "official" OCZ memory support thread at the Anandtech forums, maybe you can find something interesting there:
http://forums.anandtech.com/messageview ... erthread=y
http://forums.anandtech.com/messageview ... erthread=y
lol, this is almost creepy. I posted my problem last night, and just this morning a new guy registered and posted an almost identical problem.
http://www.bleedinedge.com/forum/showthread.php?t=19000
Anyway, I got a response from one of the support staff and it looks like I'm going to RMA it. Their customer service is impressive from everything I've seen so far.
http://www.bleedinedge.com/forum/showthread.php?t=19000
Anyway, I got a response from one of the support staff and it looks like I'm going to RMA it. Their customer service is impressive from everything I've seen so far.