NEW MMO-Champion CPU Overclocking Leaderboards

[Rixis]

living in province of quebec you silly, or using hydro-quebec powa?

in england you can get something called economoy 7, where your electric costs about a third per unit

(so my normal units are just over 16pence, but between midnight and 7am it's only 5.6 or so pence per unit)

therefore, being nocturnal, i'm awake and using my computer at night when electric is cheaper (using computer = not willing to sit with p95 running)

and when i'm sleeping is during the day, where the electric costs 3 times more than my normal (nighttime cheaper) rate

woo for nocturnality

but i'll stop derailing now :P

oh and i love how the spambot 2 posts up stole my post without even quoting me

[Ghâzh]

So you would run Prime95 instead....

I've already spoken with a couple other overclockers, including a moderator, and I think this should be fine for most people. It is certainly better than 0 rules and more or less allowing you to post any OC you want without any real proof of its stability, because that was the problem with the previous overclocking leaderboards thread. Like I said though, this is currently very open to discussion as I just made this and we can easily add/change things right now.

I'm just looking for ways to show some consistence. If there's one guy running prime 95, other IBT and the last F@H, we have three different results about the stability of their system and the three aren't comparable with each other.

I'm not saying that prime 95 is the only way to stress the processor. I'm saying you should pick one so we could have more consitent results. Prime 95 just being one good way to do it.

[Sephiracle]

As we've discussed in a few places, with the Sandy Bridge line, P95 seems to be superior in stability tests. AMD and older Intels are better with IBT, and I think a 10 pass with IBT is a bit on the low end, I would probably look at 20-25 passes with Max memory.

[DeltrusDisc]

Here's mine for the record:

Name: Deltrus
Processor: Intel i5-2500K
Clock: 4.4GHz
Voltage: 1.32v
Cooler: Zalman CNPS 9900ALED
Screenshot Link: http://i.imgur.com/LQVEb.jpg
Valid CPU-Z Link: http://valid.canardpc.com/show_oc.php?id=2092453

Updating thread now since I just got up.

---------- Post added 2011-11-11 at 02:59 PM ----------

As we've discussed in a few places, with the Sandy Bridge line, P95 seems to be superior in stability tests. AMD and older Intels are better with IBT, and I think a 10 pass with IBT is a bit on the low end, I would probably look at 20-25 passes with Max memory.

Xuvial was telling me one test should be enough. x_X lol

---------- Post added 2011-11-11 at 03:02 PM ----------

I'm just looking for ways to show some consistence. If there's one guy running prime 95, other IBT and the last F@H, we have three different results about the stability of their system and the three aren't comparable with each other.

I'm not saying that prime 95 is the only way to stress the processor. I'm saying you should pick one so we could have more consitent results. Prime 95 just being one good way to do it.

Honestly, I think right now it comes down to if the CPU is being pushed to 100% usage on all four cores, which, from what I've gathered through using all 3 tests myself, each of these do. IBT pushes out more heat, so you get a shorter test in comparison, F@H requires a pretty good period of time (at least for me) to finish one SMP WU, and Prime95 can take quite a while if you want to do the entire test as well. I think they all work fairly well, just trying to give people options, since some people might not be comfortable with running P95 overnight or while they're away at work (who knows) but they'd have the option to just run a quick session of IBT instead.

Debate is still open, however!

[Sephiracle]

Xuvial was telling me one test should be enough. x_X lol[COLOR="red"]

I don't mean doing multiple tests. I just think that if you're running a Sandy Bridge you should be using P95, and not IBT.

[DeltrusDisc]

So do you want me to tell Sandybridge people to do P95, 20 tests of IBT max, or F@H? >_<

[Sephiracle]

So do you want me to tell Sandybridge people to do P95, 20 tests of IBT max, or F@H? >_<

I don't know anything about the folding crap. I think if you have a SB, P95. If you don't IBT.

[DeltrusDisc]

Omg it's not crap! >_< It's quite a CPU/GPU intensive application! AND IT IS FOR HELPING TO FIND CURES FOR HUMANITY!!

[Lyaral]

I look forward to an updated ss, that's a killer voltage though, very nice! My 4.4GHz OC requires a 1.32vcore and 4.5GHz requires me to go up about another.05v.

http://i.imgur.com/bKULK.jpg I think that's everything covered now?

[Sephiracle]

http://i.imgur.com/bKULK.jpg I think that's everything covered now?

Yeah, and what speed? You took the pic after the test was complete and it had gone back to 1.6 lulz

[DeltrusDisc]

He was the first to post originally, he was just missing CPU-ID, but I'll take his, since he has a valid CPU-Z link.

[Ghâzh]

Honestly, I think right now it comes down to if the CPU is being pushed to 100% usage on all four cores, which, from what I've gathered through using all 3 tests myself, each of these do. IBT pushes out more heat, so you get a shorter test in comparison, F@H requires a pretty good period of time (at least for me) to finish one SMP WU, and Prime95 can take quite a while if you want to do the entire test as well. I think they all work fairly well, just trying to give people options, since some people might not be comfortable with running P95 overnight or while they're away at work (who knows) but they'd have the option to just run a quick session of IBT instead.

I think you're after a good cause but your reasoning is still somewhat flawed. The point is that sometimes you can pass an infinate amount of IBT runs and still fail in prime 95 and that could potentially be exploited. Like, why would you even want to run 8 hours of prime 95 instead of 10 runs of IBT with sandy bridge if you were aiming to score well?

[DeltrusDisc]

Inb4 I get flamed for this decision....

All posts prior to this one are grandfathered in no matter what stress-test they used.

Intel Sandybridge is only open to be stress-tested by either Folding@Home or Prime95.

As it stands right now, P95, F@H, and IBT are all fine for any other CPUs.

[Cyanotical]

there is not one true test for stability, all you are trying to prove is that your clock is usable daily, and not just a boot up once thing, prime95, multiple IBT and folding are all different ways to do it, as long as you are proving that your cpu is stable enough to run at 100% for long periods of time

what i would change the requirement to is a program that can consistently push your cpu to 100% for extended periods of time, and then list examples such as prime95 and IBT, but dont limit qualification to them

[Asera]

Maybe now would be a good time to restore my overclock, especially with winter approaching.

Although I'm liking my super low idle power draw right now... xD

[DeltrusDisc]

If I went that route, I'd have to ask for the task manager with CPU load shown as well in the screen shots.

[Cyanotical]

If I went that route, I'd have to ask for the task manager with CPU load shown as well in the screen shots.

you asked?
http://i.imgur.com/Xo0M4.jpg

[DeltrusDisc]

Holy shit you have your GPU processor clocked high! O_O Newegg showing most 590s have the core clocks in the 600 range.... is that supposed to be double or do you just have a truly killer OC?

And sigh@you for pressuring me to change the rules MOAR.

[Cyanotical]

i can only assume that it is combining each GPU clock, they are in fact 630Mhz, but two gpus per card, so...630 x 2 = i need to update the evga tool

[DeltrusDisc]

That EVGA tool looks badass, either way.

Btw check your PMs!