GTX 680 PCIe 3.0 & Rampage Mobo

[Arne_Saknussemm]

How many of you with Rampage 4 Extremes have had no problems running your GTX 680 at PCIe 3.0 via the "hack"

Negative experience is often up on the forum but not so much when something works so I thought it would make an interesting stat. Plus I´ve just ordered a DCII top and wondered what the chances were....:)


EDIT:

Manual hack: http://www.evga.com/forums/tm.aspx?&m=1607292&mpage=1

Nvidia force enable: http://nvidia.custhelp.com/app/answers/detail/a_id/3135/session/L3RpbWUvMTM0MDIyMzU2OC9zaWQvaDEzbE45...

EDIT 2: One person has found that running two cards initially broke 3.0 but changing the order of the cards in the Motherboard slots got it back up and running. Suspicion is that some GPU chips may better handle the extra load that the first card has to bear in the relationship with the CPU.....

[Arne_Saknussemm]

I see you are at a constant 1.35 vcore are you running manual voltage for an overclock? Are you running XMP profile. Have you ever had any trouble from the system before?

Maybe you could try to run 3.0 at default BIOS settings and then run XMP and 3.0 and then run overclock and 3.0 not XMP and find out if any one of them is influencing........ or philosophically assume that for one card the difference is very little.....

Nodens wrote:
there's nothing you can do though.

Hey Nodens...is that right....nothing to be done. No tweaks that affect PCIe signalling etc.? If the CPU is not running the show properly it's kind of all or nothing? Is this and IMC strength weakness related?

[PSoares]

PSoares can you verify that the issue only happens with GEN3 on? Like running the benchmark with GEN3 and GEN2 a couple times on each so we can say for sure if it's GEN3 related? If it is there's nothing you can do though.

Nodens,
how can I revert to GEN2?
Sorry but I´m very amateur in this field.

I see you are at a constant 1.35 vcore are you running manual voltage for an overclock? Are you running XMP profile. Have you ever had any trouble from the system before?

Arne Saknussemm,
I have a little OC, from "Easy overclock settings - from ROG".
I will trying without OC.

Thanks for helping! 😉

[Nodens]

Well in theory overclocking could affect it I guess. It should always be tested at stock settings to be sure. But imho it's not worth it to lessen your overclock just to enable GEN3 which currently will offer you nothing in terms of real world performance. Your higher overclocking offers better performance anyhow though. So the trade off is not really worth it even if you could get it stable that way, no? That's why I said there's nothing you can do, as a blanket statement. If it doesn't work when you just enable it, then it's not worth it doing anything that would lessen your performance to make it work.

Psoares, use "force-enable-gen3.exe -revert" to revert the change and reboot. If that doesn't work for any reason, you need to manually delete the "RMPcieLinkSpeed" variable from the registry.

[PSoares]

I have tried with&without OC and the problem was there.
Disabling GEN3 the game runs smoothly!

[Arne_Saknussemm]

Hmm so stay on Gen 2.0 best then!

I wondered about your overclock because 1.35 volts is quite a lot and should on average (of course each CPU is different) get you more than a little OC. Also if you run constant voltage through your CPU it will not live as long as one that lets Speedstep work and lower voltage.

Maybe try a negative offset of -0.005 and see if that lets your voltage come down?

[PSoares]

I will try it!

[Arne_Saknussemm]

Hmm I didn't know you had it working on one but not on two. Weird. You should swap out the card that does work from the lane it works in and try the other card. If it doesn't work it may be the second card and not the board/CPU.

If you do play with those other settings let us know the results I would be interested in knowing if they have an effect.

🙂

EDIT: actually my cards do throttle up and down 1.1 2.0 3.0 and the link state power management is off in performance power plan......

[Nodens]

It makes sense that it only works on one card. Second card adds additional strain, revealing the incompatibility of the CPU. Just forget GEN3 until we can upgrade to IB-E...it's not like you'll be losing any speed...GEN2 is not currently saturated so you won't see any real life difference by GEN3 that is not in the margin of error.

[Lettuce]

Well I tried a few things;

The 2nd card could do 3.0 in any slot as long as the 1st card wasn't present in the system.

So I then tried swapping the cards, the one I had in the top slot, now in slot 3 and the bottom one in slot one...

Everything works now. What the f**k?

Both cards are stable for ~6 hours of stress testing, both at 3.0 speed.

The thing is, there is no reason I can think of that swapping the cards positions should actually have fixed it. But I guess I won't complain and just enjoy it.

I experimented with a few other things, and a few settings in bios but I reverted most of them and it still works.

If anyone else is having issues, I'd suggest trying swapping your cards positions, along with disabling power saving on the PCIE link in windows power management, and setting PCIE spread spectrum to disabled in BIOS.

Thanks for your input Arne and Nodens you pointed me in the right direction.

[Arne_Saknussemm]

Cool bananas! That's a good result...but yes, a strange one. I wonder how that comes about. Still I guess the pcie lanes link to the processor in a fixed wayif the processor is having signal issues with one card over certain lanes why not try a different card or different lanes.
All's well as ends well! 🙂