Memtest error and then not anymore? Could it happen again?

[chriscrows]

Hey,

I just built a new system:

Intel Core i7-6700K 4000 1151
Asus Z170-PRO
Corsair Vengeance® LPX 32GB (4x8GB) DDR4 DRAM 2666MHz C15(CMK32GX4M4A2666C15)
Nvidia GTX 1070 Founders Edition Asus D5
2 x Samsung SSDs one 250 and one 120 GB (I had them already)
WDC 1TB HDD (I had it already)

Corsair RM750i 750W ATX24
Cooler Master Hyper 212 Evo

After I installed Windows 10 and the rest of my stuff I got at one point a BAD_POOL_HEADER. From some research on the internet some people mentioned that it might be the RAM.

I got memtest and ran it. In the first few seconds it gave me errors. I removed 2 memory modules and I got no more errors. So naturally I thought one module would have a problem. Then I removed the "good modules" and inserted the "bad ones". No error.
I also tried with 3 modules but then I get only a black screen(might be because they work 2 by 2? I'm no expert)
http://imgur.com/kgTyY3F

I put all the 4 modules back and after reading more on the internet on here some people suggested to increase the RAM Voltage. From 1.2V I went to 1.35V but still got the errors.

Then I increased the CPU voltage from 1.280V to 1.35V and I got no more errors. BUT here is where it starts to go into X-Files territory. After putting the CPU back to auto(1.280V) the error disappeared in memtest.

http://imgur.com/1ArKRPY

So now my question is...where did the error go? Could it be fix temporary or .....I have no idea...

I have to say the Memtest gave the error in the first seconds. Now it's running for 43 minutes and no error.

Thanks and sorry for the long text, I wanted to be as concise as possible.

 

[kanewolf]

MEMTEST needs to run OVERNIGHT or 24hr with zero errors to be considered stable in my book.