Rampage IV Formula + Gskill XMP Problem

[Fong]

Hi,

I've just built my first intel system using the RIVF and F3-1866CQF-16GAB from G Skill. This memory kit is rated at 1866 9-10-9 1.5v. I noticed that with everything on auto, it will run at 1600 cl11. However, if I switch to XMP, I will start to get errors when I run LinX at around the 3rd run. Everything else is stock. I have tried setting rampage tweak to mode 1 and vccsa to 1.25v as per G Skill's instructions for the RIVE (even though I own the RIVF) but I still get errors when I run LinX. What can I do? BIOS is also updated to the latest.

Thanks

[HiVizMan]

What I would suggest is downloading memtest ver 4.2 (I use the USB version) and test to see if the issue is with your ram or the motheboard slots.

Test each stick individually at xmp for at least 2 completed cycles.

Then take a stick that has passed and test each slot in the same way.

Once done, you can then add on stick at a time and run the test, till all the sticks are in system.

That will pretty much identify where the problem is from a ram perspective.

*******


You have not mentioned if you are overclocking your CPU at all and I wonder about that.

[Fong]

The system is at stock with XMP. That might be a bit difficult since I'm running the CPU on air and the heatsink covers the memory slots but i'll try. If I leave it on auto 1600mhz cl11 LinX runs fine. Also, I noticed that on auto my 3930k will turbo to 3.5 while if I set it to XMP it will turbo to 3.8ghz? Is that right?

[HiVizMan]

Turbo should not alter with XMP and you will find that the difference is rather to do with what background services are using resource. Turbo max is single core. If you have a couple of back ground stuff going on then only two cores will ramp up. And that is 3.5

As to testing metest that is of course your choice. 🙂

[Fong]

But when I run LinX at auto it will turbo to 3.5 while using XMP it will turbo to 3.8. Wouldn't LinX be stressing the cores in the sae way? How come there is a difference in turbo? Sorry bit of a beginner...

I will try memtest and also adjusting the dram voltage from 1.5v and report back. Thanks

[Fong]

I just ran memtest for one pass with no errors. Is one pass enough to say that it's not the memory? As I said before, with everything auto the cu will turbo to 3.5ghz when I run LinX but will turbo to 3.8ghz when I use XMP. Is this normal? What should I do next? Thanks

[8_Pack]

Linx is a test based on IBT which is for the CPU. How are you sure this is a RAM problem it may be CPU orientated. I would do as Hiviz suggests and try isolate the problem by running memtest. This will allow you to eliminate the RAM from the equation. Remember VTT, PLL and clockgen filter also help alot with RAM stability. You could also try bumping the RAM voltage to 1.55 or 1.6v which wont cause any problems with those sticks and may eliminate the instability.

[HiVizMan]

Did you run memtest at the XMP profile?

[Fong]

Yes memtest was done with XMP loaded. I think I may know what the problem is. I did some googling and apparently ASUS sets the the turbo ratio to 'by all cores' when XMP is loaded. This would explain why I'm seeing 3.8ghz even when LinX loads all cores. I'm still doing another pass of memtest so I can't go into BIOS to confirm. Perhaps my CPU is just unstable at this turbo setting?

[HiVizMan]

If you have passed two cycles of memtest at the XMP profile then you can safely assume that your ram is stable at XMP.