• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

i7 970 help

Gio

Weaksauce
Joined
Nov 25, 2005
Messages
68
Hey guys I need some help. Got myself up and running with a i7 970 with a Rampage 3 Gene, and 6GB of Kingston Hyper X 1600. Cooling it with a H70 w/ GT15's. Not new to overclocking and I've been reading post after post and guide after guide with this i7 method of overclocking and my head is starting to spin! :confused: Couple questions I have is:

1: How important or reliable more importantly is prime 95? I have the latest 64 bit version and for the life of me it is never stable under prime. I use it as a last resort test but haven't cared much for it in the past. I use IBT Very High 40 passes after i get a solid 10 passes on standard. (set manually to 12 threads) but prime fails torture tests and temps are reasonable. I currently have it set to 1.25v CPU and QPI, 20*200 BCLK. Seems very stable other than prime. I have even loaded into windows as high as 4.6GHz just for giggles and passed standard IBT.

2. To utilize the 25 turbo multi does turbo boost need to be enable or is there another setting? I can select 25 and boot with it but I notice it doesn't stay consistantly at 25, it drops down to 24 then back to 25.

Lastly any other advice for overclocking this thing?

Thanks a bunch :D
 
I'll take a shot at #1. Generally, any Prime95 instability reflects potential overall system instability. But, there are exceptions to every rule; maybe your system is an exception based on the IBT results. However, which stress test are you running? Small, large or blend? How long does it go before it fails? Does it always fail on a specific size/range of FFT? Stable to other stress/gaming? Will it pass Prime95 at stock settings?
 
Take some screen shots of your BIOS template so others can help you better and run P95 on stock settings/OC.
 
I was running torture test for Prime. My method is 5 quick passes of IBT at 12 threads on standard. Then 20 passes at the same setting. Then I bump it to Very High instead of standard and run 5 passes there then 20-40. Once all checks out i run Prime torture test (blend I believe it is) and go to bed. So unfortunately I don't know exactly when it fails but I wake up with a BSOD laughing at me.

For the BIOS template that is an awesome idea but how do I do that ? lol sorry for sounding like a n00b
 
As far as I know the only way to post BIOS screens is the old-fashioned way- take some pics of your monitor with the digital camera at each screen.

You might switch to the small FFT torture test which should only test the CPU. See if you can pass that, then try the Large FFT test which focuses on RAM. The blend will throw large and small FFTs at your computer testing both CPU and RAM and making it harder to figure out middle of the night crashes.
 
This is really strange, I have the opposite experience, If Prime fails generally, my IBT fails really quickly. However I have never seen Prime fail if my IBT runs very high flawlessly 20 times. You seem to have an amazing chip. I can't even touch 4.2GHz with my 980X.
 
I'm going to write down the bios options and settings tonight and post them up as some settings I'm sure could be tweaked. Currently after messing around last night the CPU is at 20*200, 1.256v in bios, QPI 1.25v manually set the memory to 1603 and 7200 something for the quickpath. Loadline is enabled for the cpu and QPI. In windows the cpu will go up to 1.25/.1.26 underload which is good and the QPI will go anywhere from 1.25-1.3v underload. Everything else is on auto.

I even managed to get 4.2GHz somewhat stable under IBT but I stopped testing because the core temps got as high as 89*c and this is with the H70 push/pull. Stressed it stock and the cores never got higher that 53*C and the CPU was running at 1.09v. From my past experience I should have some headroom but I dont know if im doing something wrong or forgetting something.
 
It could also just be some Prime95 issue unrelated to your PC. From my OC research it seems like a lot of folks on the forum don't even use it anymore now that IBT is around (maybe it would even fail for them too?). Just like IBT and OCCT etc, Prime95 is a tool and not the be-all end-all for determining OC stability. Obviously it should work, but maybe after a bit more testing if you can't figure it out, maybe you just shrug your shoulders and move on.

I ran my E6400 at 3.2 GHz for a year without any *apparent* issues (certainly no crashes that I couldn't explain) despite usually getting a worker error within the 1st hour of running Prime95. Maybe (probably) I was rolling the dice, but I am a relatively light user and the PC was always stable to gaming, browsing, etc and rarely under a constant load for more than 2 hours. Maybe that critical error was only another cycle away, dunno.

I guess, do some more testing but also consider that this could just be an anomaly.
 
Yeah for now I am ignoring prime95 unless otherwise directed otherwise. What really gripes me right now is that 2 of my buddies have the same chip on Sabertooth boards, both running @ 4.2 with 1.35 and 1.375v CPU, both under air coolers, True 120 and Silver Arrow and they are getting lower temps low 70's max. The only thing that changed on my H70 from before to now is I had to change the thermal paste from stock to MX4. So I'm going to reapply and hope for the best.
 
Thought I'd update. Reapplied the MX-4 and temps dropped nicely. I currently have it running at 4.2GHz 21*200 with 1.3v. Full load temps under IBT very high test are low 70's. (74/79/64/70/73/74). Very happy with the outcome.
 
Back
Top