Memtest86 failed three times (2 hours), twice (20 minutes), and then nine times (10 hours)

michaelkm8

Honorable
Jun 28, 2015
55
0
10,630
As the title says my ram failed memtest86. For the first test I had the following:
- Mushkin redline enhanced 4x4 2133 MHz (16GB) @ 2133 MHz, 9-11-10-28, 1.65 V

For the second and third test:
- Mushkin redline enhanced 4x4 2133 MHz (16GB) @ 2133 MHz, 9-11-11-28, 1.655 V

Should I just RMA it, got it Friday, or should I adjust the settings to make it stable? This is the first time i've dealt with testing computer hardware so I don't know what to do after something fails. Thanks again for all the help.

Update:
- Running Memtest86 V 4.3.
- Mushkin Enhanced Redline 8GB (2 x 4GB) 240-Pin DDR3 SDRAM DDR3 2133 (PC3 17000) Desktop Memory Model 996996
 

michaelkm8

Honorable
Jun 28, 2015
55
0
10,630


Thanks!
 

michaelkm8

Honorable
Jun 28, 2015
55
0
10,630


I have profile 1 activated for XMP and am running the ram @ 2133 MHz with custom timings 9-11-11-28 1.655 V. The timings are from the manufacturers specifications.
 

michaelkm8

Honorable
Jun 28, 2015
55
0
10,630


I did memtest86 V 6 and it only gave one error in test 13.
 

michaelkm8

Honorable
Jun 28, 2015
55
0
10,630


I5-4690K
Gigabyte Gaming GT
Gigabyte GTX 970
Mushkin Redline enhanced 4x4 16GB 2133 MHz
Samsung 850 EVO 500GB
Corsair AX850 80+ Gold PSU
 

Tradesman1

Legenda in Aeternum
Try 4-5 passes on each stick individually, testing with multiple sticks can create (false) errors, test 13 is the Hammer test which has produced errors for many, though the systems themselves haven't really been effected negatively. If errors persist can try adding a bit of voltage to the MC (memory controller) maybe + 0.06 to, believe it's the CPUVTT or DDRVTT voltage
 

michaelkm8

Honorable
Jun 28, 2015
55
0
10,630


Will do.
 

michaelkm8

Honorable
Jun 28, 2015
55
0
10,630
So I contacted the tech department at Mushkin and told them the ram failed with 22 single bit errors in test 13 and specified it was the hammer test. They emailed me a few minutes later apologizing and sent me a shipping label to RMA them. I only asked if this would be a problem and didn't ask for a refund or replacements, so either they know it would be a problem or they didn't want to bother with it.
 
same here - but did you run the memtest86+ to see if it gave errors as well ? that f.a.q from the passmark memtest seems to state its a known issue in some ways [hammer test 13 ] ?? so it may not be your memory at all ??

it not going to hurt anything but down time to do the rma and it may resolve as well ?? time will tell on that

good luck
 

michaelkm8

Honorable
Jun 28, 2015
55
0
10,630
So I talked to a tech at Mushkin about the hammertest and he said that they usually were false positives if only the hammertest were receiving errors. I downloaded Memtest86+ and ran it for almost 14 hours and it ran through 6 full tests and I received zero errors. Again Memtest86 only gave me errors in test 13 so i'm inclined to agree with Mushkin about the reliability of the ram modules I bought. I am going to read the paper about the hammer test maybe this weekend and i'll update what I get from it, but here are my initial thoughts about the bit flipping. First something of this nature is obviously a quantum problem and relates to quantum tunneling or the probability of energy being located in a certain position over a certain amount of time. This being said for this effect to be realistic in terms of normal everyday use the ram would need to be heavily used so that each row was activated within a time frame that allowed it to be activated simultaneously in conjunction with a row in a position so that one un-activated row were between them. From what I have read the probability of every day use computers that do gaming, emails, 3D printing applications, and other such processes have a low probability of "hammering rows" in such a way. This leads one to understand that this effect will most likely occur under extreme conditions and by chance, although the chance will be slight. Again this is my first impression and I'll update later as the only answers I've gotten from the Memtest people are that errors are absolute, and from the Mushkin people that say they are false positives.
 

TRENDING THREADS