• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Gskill Ripjaw-z 2133mhz only in Dual channel error

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

alegelos

Member
Joined
Feb 2, 2010
Location
Uruguay
One pc i build have error when running memTest. But only when testing in Dual channel. It have 4x4gb 2133mhz cl9-11-10, volt 1.65. All tested separete with none error in memTest.

But when 2 or 4 modules in dual channel, then memTest goes with 999 errors.

I try volt at 1.7 but same problem. I am running them ar 1333mhz cl9-11-10-28


Mother Gegabyte 990xa
CPU: AMD FX-8120
PSU Corsair 850

Any idea?
 
Put the ram in single channel mode so the PC will run reliably and then download and install the program "CPU-z". Run the program and attach screenshots of tabs: "CPU", "Memory" and "SPD". To do that, first crop the images to disk with Snipping Tool found in Windows Accessories. Then click on "Go Advanced" at the bottom of any new post window. Then click on the little paperclip tool at the top of the Advanced post window. That will bring up the file browser and uploader. The rest will be obvious. You can attach up to three pics per post. You may need to set your memory timings manually. CPU-z will tell us what the timings need to be and the voltages.
 
I don't know why you trust so much in cpuz SPD tab. When you use X79 memory with XMP 1.3 then cpuz will read it wrong on AMD/990FX.
All 2133 XMP 1.3 profiles that I made myself were working as 2000 on AMD boards and under a bit different timings.
Thats why memory like Ripjaws-Z are dedicated for Intel XMP 1.3 compatible boards ( Z77/X79 ).
Thats also one of reasons why G.Skill has only X79 boards on qualified motherboards list:
http://gskill.com/products.php?index=438
 
Like I said you have there 1778 or 2000 9-12-11-29 what is wrong read from XMP 1.3 ...
Try manually 2133 9-11-10-28 , command rate 2T <- this is a must for stability, 1.65V vdimm and CPU-NB +0.10V ( if you make it work like that then it's ok , if not then try +0.15V )

other timings that you can try:
tRC 41-48 @ 2133 or 39 for < 1866
high tRFC like 180+
 
I set ram always at 1333mhz and 2T

I tried +0.1, 0.15, 0.2 on CPU-NB
Tried TRC at 48
Tried tRFC 110, 160 and 300
Tried Ram voltage 1.7 (stock 1.65)

And all same error.

Those ram in my pc in dual channel work with none error. Or in that PC in single channel also none error.

Any idea?
 
yes also 1600 and 1866. I cant get 2133 using the 4 dims. But yes i try 2133 and fail.
 
Probably noone will guarantee that this memory will work on AMD/FX setup.
You can make it work but on manual timings and I have no idea how full SPD table for this memory looks like.
You can ask G.Skill support what to do with it or if they can send you full timings list.
 
I don't know why you trust so much in cpuz SPD tab . . .

Woa, that felt a little angry!

I trust the CPU-z tab because my understanding is that it is just reporting information encoded on the memory module by the memory manufacturer. And also, I have not had it give a bum steer in my own experience. But then I don't work with Intel systems any more.
 
yes also 1600 and 1866. I cant get 2133 using the 4 dims. But yes i try 2133 and fail.

4 DIMMS is not just "dual channel"

Harder to run 4 dimms than 2. Defintiely will need to boost CPUNB and as woomack pointed out indirectly.. manually settings timings will help. Might need a buttload of TRFC since they are intel "optimized."

EDIT: just reerad thread and saw woomack also recommended high TRFC... lol. Sorry for the "repeat" post :)
 
Last edited:
In your bios screen shots you are raising CPU voltage NOT CPUNB. (directly above that voltage you changed is the CPUNB votlage)
 
Only 2 dims in Dual channel with CPU NB 1.35 are stable and no memTest error. But 9999 errors with 4 dims :S

Any idea?
 
If I were you I'd take all the ram sticks out but one and move it around to each slot and test each slot. Maybe you have a bad slot.
 
Back