G.skill F317600CL7D4GBXHD woes

[aussie-revhead]

Guys, I just received these C7 rams from memoryC but Im having great trouble getting my rig stable with them installed. Oddly enough they pass memtest up to 2200-7-10-10 but anything over around 2000MHz and my rig is unstable. I have tested 2200+2133 and both fail, so far 1600 passes and 1866 looks to be passing too, so is this a compatiibility issue with my P8P67 EVO?

Testing was done with cpu overclocked at various levels but as I found it was unstable I dropped it back further and further until it reached stock clocks and it was still unstable.

With Kingston HyperX 2000 rams I have no issues up to my bclk limit on 1866 multi, its definately something to do with these rams. Or is it to do with the higher multi as I have never used this with my lower spec rams?

Any help would be greatly appreciated.

:up:

[GSKILL TECH]

Do you have the latest BIOS? What settings are you using? Post some pictures and we can take a look to see what may be holding you back.

Thank you
GSKILL TECH

[Hondacity]

are you just overclocking your ram? or is this a combination with cpu overclock?

bios settings please?

just helping out :D

[aussie-revhead]

Thanks guys, I just cleared bios so all at default, change one setting only - AI Overclock to XMP profile , this sets the ram to 2133-7-10-10, boot into windows and run prime, one thread fails within a few seconds, reset prime and try again and the same happened. Next try lasted a little longer but still failed, see screenshot.

:shrug:

Edit1: OK screenshot 2 shows its stable enough at 1866-9-9-9, these are default settings with only ram speed set in bios and vdimm set to 1.65, all other bios settings at default. (will add screenie later, its too big by 19kb ...)

Edit2: Instead, screenshot 2 shows stability at 1866-8-9-8-24, as above- all bios settings at default other than ram speed, voltage aand timings.

:up:

[aussie-revhead]

OK, hondacity prescribed a bunch of tests via PM as follows but this ram fails the lot:


i check your rams and they're suppose to work perfectly...with sandy bridge

just to make sure everything is ok.

test1. raise vcore 1.25volts,,, and using xmp profile..if this doesn't help go to test2..while keeping vcore @1.25v
test2. raise vdimm 1.68v, then run your test..if failure still go to test3. keep vcore and vdimm raised.

test3. raise vccsa to 1.1v...then run your test...if failed... keep everything raised.. go to test 4.

test4. this is where you test vpll ...stock is 1.8v
you have to test 1.65v....1.67v 1.7v 1.73v....untill you fix the problem...if not...



I tested vdimm up to 1.7v and PLL from 1.65 to 2.0 but it still fails.

:shakes:

[GSKILL TECH]

Did you adjust VCCIO any? What does it currently show for that value?

Thank you
GSKILL TECH

[aussie-revhead]

No I only adjusted that which I was told to, particularly during this diagnostic phase. Currently VCCIO is 1.05v (default), what range should I test?

:up:

[aussie-revhead]

OK I have found that one stick primes fine for hours at 2200-7-10-10-27 (XMP profile with 103.2 BCLK) while the other fails in minutes at 2133-7-10-10 (XMP) , so one stick is stuffed, how do I go about replacing a stick I bought from MemoryC ? Do I have to send it back to USA or is there someone in Australia that can take care of the RMA?

Thank you.

[GSKILL TECH]

Contact MemoryC to see what they can do. Through G.Skill, you will need to RMA to our Taiwan office.

http://www.gskill.com/rma.php

Thank you
GSKILL TECH

[aussie-revhead]

The rams are on their way back to Ireland for RMA .... fingers crossed....

:up:

[Leeghoofd]

Wasn't Taiwan the cheaper option for RMA ?

[aussie-revhead]

It wasnt dear to post, thats the rma proceedure with memoryC anyway so I must obey right?