Are results more accurate in 1 core vs multicore testing?

[gtsamour]

Thank you David.

I already submited a ticket for RMA at Corsair with pictures of the memtest86 showing the error and it has been aproved for replacement.

[David (PassMark)]

Second pair sound like they are bad, yes I would try and get them replaced.

[gtsamour]

So should i take the error report as accurate and contact corsair for replacement?

Im in the middle of testing the two pairs individually right now.
The first pair which i bought two days ago reported no errors but gave the high frequency note three times (multicore test).


The second pair reported an error in the second pass. I renun the test to see if the error comes up again. It came up again.


Do you think i should send this pair for rma?

So multicore testing is more accurate in revealing errors due to more stress?

I doubt its an overheating problem, i use watercooling on the cpu and the ram is not overclocker or overvoltaged.

[David (PassMark)]

I don't think it is a question of accuracy, but rather question of test time. e.g 1 error every 12 hours of operation. If you ran the single core test several times it probably would have eventually also got an error I suspect.

Running multiple cores gets more testing done in the same period of time, so it does increase the load on the RAM and the temperature of the system. So if the fault condition is temperature sensitive running multiple core might expose the fault.

[gtsamour]

Im a bit buffled... I tested my 16gb of ddr3 corsair vengeance lp ram using the default settings (1 core testing) and after the four passes i got no error. Then i thought about testing using all four cores in parallel.... I got no errors during the first three passes but on the fourth one i got two errors.

My motherboard is asus maximus vi hero running the latest bios, the system isnt overclocked and my ram is 1833mhz xmp 1.3 running at that speed and its standard voltage which is 1.5v. My cpu is i5 6870k


I use four dimms of the exact same brand and type but from two different packages. They are placed correctly though which is two dimms of the same pack in the same coloured slots.


Any ideas as to why this is happening? Is mukticore testing more accurate than single core testing thus finding something the single core test missed? I run the uefi version of memtest86 v6.3.0