Memtest86 reports 30000+ errors so far...safe to say I should RMA my RAM?

[Is]

No, but seriously, is there any other reason this would happen at stock timings, FSB speeds, and voltages? Any other reason at all?

 

[Mercurien]

Some Corsair ram likes a little extra voltage.

 

[Is]

That's a very good tip, and I do have Corsair RAM, but it did not help a bit. All of the tests fail miserably.

 

[us3rnotfound]

does your computer work in windows? You never said whether or not you are actually running an os. If so, does it crash a lot. We need more information

 

[Is]

Ah, well, it does work in windows, and it does crash a lot. I actually tried 3 installations of windows, and I tried 98SE and XP because I never even expected the RAM.

I've got:

Antec 350W PSU
P4 2.4C
Corsair XMS 3700 platinum (1gb)
ASUS P4P800 Deluxe
Sapphire Radeon 9800 Pro 128mb
Sony DVD/CD-RW drive

I had pretty much tried everything else in terms of installing/uninstalling software/hardware before this, to no avail of course.

 

[Yourself]

30,000 errors.....it's defective...no question.

Self

 

[Corporate Thug]

R-M-A

 

[NogginBoink]

Either the ram or the motherboard is bad.

 

[JustAnAverageGuy]

R-M-A and R-A-M are spelled with the same letters... Ever notice that?

 

[Boonesmi]

Originally posted by: shady06
R-M-A

what he said ^^

 

[InlineFive]

Originally posted by: Boonesmi

Originally posted by: shady06
R-M-A

what he said ^^

Seriously, R-M-A! Some people on here are mad when they get 10 errors! :Q

 

[ZL1]

Originally posted by: PorBleemo

Originally posted by: Boonesmi

Originally posted by: shady06
R-M-A

what he said ^^

Seriously, R-M-A! Some people on here are mad when they get 10 errors! :Q

thats true, Id be mad if I got 1 error actually

but consider this, the board might be the problem, or you just didnt insert it well (it does happen)
try removing and inserting again then try to put some other ram in to see if the board is aok, if the first one didnt help and the second one passed RMA the ram

D

 

[Is]

Unfortunately, the only other RAM I have is PC100, I don't think that will work. I'm testing each module individually (it's twinX doncha know), both report errors. It is sort of weird that two modules in one package would both be bad, but I suppose they would be from the same batch.

 

[jamison]

According to Mushkin, one error in Memtest86 is justification to RMA.

 

[ZL1]

Originally posted by: Is
Unfortunately, the only other RAM I have is PC100, I don't think that will work. I'm testing each module individually (it's twinX doncha know), both report errors. It is sort of weird that two modules in one package would both be bad, but I suppose they would be from the same batch.

the thing that they both report errors makes me think its the board (but thats me)
got maybe a friend with a board that will support your ram ? you could check your ram on his board


D

 

[ZL1]

Originally posted by: jamison
According to Mushkin, one error in Memtest86 is justification to RMA.

and I completly agree, that one single error could cause your system to crash
+ I dont want no errors if I pay $200 for the damn thing


D

 

[Is]

Originally posted by: ZL1

Originally posted by: Is
Unfortunately, the only other RAM I have is PC100, I don't think that will work. I'm testing each module individually (it's twinX doncha know), both report errors. It is sort of weird that two modules in one package would both be bad, but I suppose they would be from the same batch.

the thing that they both report errors makes me think its the board (but thats me)
got maybe a friend with a board that will support your ram ? you could check your ram on his board


D

Yeah, you bring up a good point, and if at all possible I will test out on another system. But I have tested each DIMM individually 3x, and I have found a consistent pattern. One DIMM reports between 50-100 errors per pass, the other reports several hundred. One would think that both would produce roughly equal amounts of errors if something on the board were at fault. That's out of the same slot each time, BTW.

 

[ZL1]

Originally posted by: Is

Originally posted by: ZL1

Originally posted by: Is
Unfortunately, the only other RAM I have is PC100, I don't think that will work. I'm testing each module individually (it's twinX doncha know), both report errors. It is sort of weird that two modules in one package would both be bad, but I suppose they would be from the same batch.

the thing that they both report errors makes me think its the board (but thats me)
got maybe a friend with a board that will support your ram ? you could check your ram on his board


D

Yeah, you bring up a good point, and if at all possible I will test out on another system. But I have tested each DIMM individually 3x, and I have found a consistent pattern. One DIMM reports between 50-100 errors per pass, the other reports several hundred. One would think that both would produce roughly equal amounts of errors if something on the board were at fault. That's out of the same slot each time, BTW.

yes that is interesting indeed, your best shot would be trying to test a different memory module on your PC or your ram on someone elses, but if you cant then rma the ram (since its cheaper to ship then the mobo )


D

 

[flexy]

i would RMA if i had only ONE error.

If you have "only" ONE error then your ram is defective. (i am assuming a correct FSB, no mobo probs etc.).
Defective Ram==RMA, no matter whether it's 1 error, 10, or 30000+

greets

 

[Nebor]

Try loosening the timings and setting 2.8v VDIMM.

 

[Mark R]

Is there a pattern to the errors in memtest? Do the errors only appear during a specific test, or are only 1 or 2 tests affected?

If the errors are just random, or they appear only in a few tests, then it strongly suggests a defective motherboard. I've found that NForce2 based mobos are VERY fussy about RAM, and some manufacturers have rather lax quality control. I recently built a batch of PCs for friends and family specificially to make faultfinding easier - all the motherboards had this fault.

 

[Sunner]

Just FYI I had a comp at work(my comp actually, got me a new one, yay!) where I got a crapload of errors in Memtest, but in the end, it turned out to be the mobo, I tried a whole bunch of sticks in various configurations, all gave loads of errors.

 

[theplanb]

I'm just curous. Is it because of the fundamental design flaw of the mobo, or just one instant with defective mobo.. if it is indeed mobo's fault

 

[Is]

Originally posted by: Mark R
Is there a pattern to the errors in memtest? Do the errors only appear during a specific test, or are only 1 or 2 tests affected?

If the errors are just random, or they appear only in a few tests, then it strongly suggests a defective motherboard. I've found that NForce2 based mobos are VERY fussy about RAM, and some manufacturers have rather lax quality control. I recently built a batch of PCs for friends and family specificially to make faultfinding easier - all the motherboards had this fault.

This is an ASUS P4P800 Deluxe. It's got the intel 865PE chipset. I would hope that ASUS has decent quality control. Maybe they don't, or maybe I got unlucky.

There is somewhat of a pattern. The errors usually start to occur at test 4, and continue for almost every test after that, although if you give it a few passes eventually the errors will occur on all tests. I have found a pattern to where the errors occur, too. With both DIMMs installed, the errors occur around the 450-510mb address range, and also the 960-1020Mb range. But with only one DIMM, this shifts and the errors occur in the 200-250Mb and 450-510Mb range. I don't know what that means, if anything. I can say with certainty that one DIMM produces many times more errors than the other.

I will try to test the RAM out on another system, but most of my friends aren't exactly l33t. I have a coworker who is into computers a lot, though. I'll try my best to get ahold of him.

 

[Is]

Originally posted by: Nebor
Try loosening the timings and setting 2.8v VDIMM.

Well, I did crank up the VDIMM, I tried everything up to 2.85 actually. There is no way I'm relaxing the timings, though. This stuff is already rated 3-4-4-8 (yeah, I know - I got a decent deal on it, when I have cash I want to replace it with that sweet OCZ 3500 that was just reviewed here), and if it can't run at that I don't care if it's good otherwise.