*The Official Tech-Forums Super Pi 1M Rankings*

Re: Tech Forums Official SuperPi 1M Top List

...except I had the same settings...

Again, like I said before, the only reason I don't use 3.6 or 3.7 as a 24/7 overclock is because of temps. So, in this case, better cooling would equal higher stable overclock. That 51mhz increase just shows that when the temps go down even a couple degrees, I was able to push slightly farther, hence a better cooler like a core contact (which can net around 20c+ less temps at a full load) would definitely help getting a nice stable overclock at higher clocks.

I literally might have a 3-5 degree difference from the move...so you see where I'm going with this.

Besides, if you read my post, I was mentioning that I would be breaking 4ghz as a max OC, never did I say ANYTHING about a max STABLE oc. This is a benchmarking thread after all.
 
Re: Tech Forums Official SuperPi 1M Top List

Actually there are a lot of things that could have effected it when you moved. You could have cleaner power where you are now for example. I think you are over stating the effect of a few degrees. CPUs aren't the sensitive.
 
Re: Tech Forums Official SuperPi 1M Top List

Somehow I doubt there is cleaner power in this place. You don't live here, you don't know.

Think whatever you want, this argument is pointless. I'm just giving my results. And seeing as when I was in my old place I had to really fight for 3.888ghz with all crazy settings, and getting 3.9ghz here was as easy as upping the multiplier with no other tweaks besides vcore, I figured that would be enough proof that a "few degrees" did in fact help.

That's my results. Plain and simple.
 
Re: Tech Forums Official SuperPi 1M Top List

When your air cooling a single degree can make a difference in a stable overclock. I'm sure this will cause an uproar, but thermal paste can also make a big difference. Sometimes that's all it takes to get that extra few Mhz and pass a run of Prime95.

My point being that 5 degrees make more of a difference for air cooling and less of a difference for water and extreme cooling. Which is the exact opposite of what was posted earlier.

Case in point. In the summer when the temps go up and my room temp raises by 5 degrees I'll have to lower my overclock for stability. With water cooling that 5 degrees of difference probably won't affect my stability and or overclock at all.
 
Re: Tech Forums Official SuperPi 1M Top List

Case in point. In the summer when the temps go up and my room temp raises by 5 degrees I'll have to lower my overclock for stability. With water cooling that 5 degrees of difference probably won't affect my stability and or overclock at all.

well yeah, WC gets you 20 degrees lower than air cooling.
 
Re: Tech Forums Official SuperPi 1M Top List

...except I had the same settings...

when I was in my old place I had to really fight for 3.888ghz with all crazy settings, and getting 3.9ghz here was as easy as upping the multiplier with no other tweaks besides vcore

:confused:

I'm reserving my opinion on the issue until I can do some controlled testing and get exact numbers for the effect a given change in temperature has on a maximum possible overclock. Either way it would be good to know exactly what to expect from temp changes.
 
Re: Tech Forums Official SuperPi 1M Top List

Sorry if I wasn't clear about the settings. The "same settings" I spoke of were with the 3.92 and 3.939 overclocks...3.9 was straight multiplier with no change but vcore. :p
 
Re: Tech Forums Official SuperPi 1M Top List

I think 20dC is a little high. Even under load my temps stay in the low 60's. I was at 4.5Ghz last nite and my temps never hit 70 while running SuperPi a Zillion times :D

Here are my idle temps :D
20degrees.jpg


My last post of the nite. Lost is about to start :D
 
Re: Tech Forums Official SuperPi 1M Top List

Sorry if I wasn't clear about the settings. The "same settings" I spoke of were with the 3.92 and 3.939 overclocks...3.9 was straight multiplier with no change but vcore. :p

No problem, I figured it was something like that.
 
Back
Top Bottom