Retail Q9650 Overclocking Thread

[The Viper]

Managed to get a 4.8 screenie...still getting freezes though...I think my NB temp is the problem.

[Linuxfan]

Managed to get a 4.8 screenie...still getting freezes though...I think my NB temp is the problem.

Good SP1M result and overclock! Stick some fans on it and give it some more Vcore, thats what I did to get a 4.85ghz SP1M...

[zoson]

I really don't want to seem like I'm crapping on anyone's results... But I've found it totally impossible to get a truely stable overclock beyond 457 fsb... I can get into windows at 500x8 and run super-pi, test prime stable for several hours... These forms of stress don't really seem to simulate actual load any more... But really loading up my computer with Linpack proves instability rather quickly... Or attempting to play crysis will crash me also... And almost invariably the computer randomly locks up(indeterminate amount of time)...

My whole system is watercooled, so I've even resorted to trying rather ridiculous voltages, all the way up and down the tree...

Anyway, I think the short of this is that I think people need to stop using superpi and prime as methods for testing, as they no longer properly emulate load situations.

[humeyboy]

That's because your Mobo ain't as good as a P45 for FSB on a Quad (Gigabyte P45 seems to be best) and probably a Dual also (any P45).

[Leeghoofd]

Anyway, I think the short of this is that I think people need to stop using superpi and prime as methods for testing, as they no longer properly emulate load situations.

these tests do not include 3d gametests but are perefctly fine to get an idea about stability... load hyperpi with an instance for each core, you will have a hard time to get it stable... 450FSB is already pretty good mate on a i780 or i790 platform with a quad... Nvidia chipsets tend to run too hot and totally crap out with quads and DDR3 makes the strain on the NB even harder... Your instability is due to the Nvidia chipset... they are far too erratic in behaviour...

[zoson]

these tests do not include 3d gametests but are perefctly fine to get an idea about stability... load hyperpi with an instance for each core, you will have a hard time to get it stable... 450FSB is already pretty good mate on a i780 or i790 platform with a quad... Nvidia chipsets tend to run too hot and totally crap out with quads and DDR3 makes the strain on the NB even harder... Your instability is due to the Nvidia chipset... they are far too erratic in behaviour...

I hear what you're saying, but with this same motherboard and my E8600 I can do 4.4GHz at 550FSB with very little voltage. I can do 4.7GHz with the same chip and lower FSB...

Is the increased FSB load of a quad really so much greater that this board would croak at 100FSB less?

My northbridge is watercooled. It peaks out around 80C with 1.85v and 500fsb at full load with the quad... With the dual it peaks out around 60C with 1.38v and 550FSB...

Anyway, what I would really like to see is one of these guys hitting ~550 fsb with a quad on one of these P45 boards is to prove what everyone is saying to me... Run 500 LinX loops and show the screenshot of it finishing.

[Zucker2k]

I hear what you're saying, but with this same motherboard and my E8600 I can do 4.4GHz at 550FSB with very little voltage. I can do 4.7GHz with the same chip and lower FSB...

Is the increased FSB load of a quad really so much greater that this board would croak at 100FSB less?

My northbridge is watercooled. It peaks out around 80C with 1.85v and 500fsb at full load with the quad... With the dual it peaks out around 60C with 1.38v and 550FSB...

Anyway, what I would really like to see is one of these guys hitting ~550 fsb with a quad on one of these P45 boards is to prove what everyone is saying to me... Run 500 LinX loops and show the screenshot of it finishing.

What do you mean by this?

[andressergio]

What do you mean by this?

I think that he says so cauz he didnt tried the GIGABYTE world yet, still on Asus so i can feel his pain

Cheers
Sergio

[andressergio]

these tests do not include 3d gametests but are perefctly fine to get an idea about stability... load hyperpi with an instance for each core, you will have a hard time to get it stable... 450FSB is already pretty good mate on a i780 or i790 platform with a quad... Nvidia chipsets tend to run too hot and totally crap out with quads and DDR3 makes the strain on the NB even harder... Your instability is due to the Nvidia chipset... they are far too erratic in behaviour...

Leeg hyperpi assigns a thread to each core so no need to do it manually inmo, and yes NVIDIA + QUAD no go agree on that

Cheers
Sergio

[zoson]

If you guys are so stable on gigabyte/p45 at these speeds, prove it by running 500 LinX loops, is what I'm saying. Because right now, I can show what you are all showing. My computer will run for 10 seconds and complete a superpi. It will also run for 8+ hours of prime95 at 520fsb, but it crashes after 30 or so LinX loops.

So once again, prove your overclocks are stable with P45. If you're so confident, there should be nothing stopping you from running LinX like I ask. Right now I think you guys arn't actually stable.

[kagula323]

hi andressergio. I see u have Gskill PI 1200 2x2GB ram. I got them too and I wonder what is the max voltage to run on them whitout active cooling and what max freguencies/ latencies can i expect from them?

[andressergio]

If you guys are so stable on gigabyte/p45 at these speeds, prove it by running 500 LinX loops, is what I'm saying. Because right now, I can show what you are all showing. My computer will run for 10 seconds and complete a superpi. It will also run for 8+ hours of prime95 at 520fsb, but it crashes after 30 or so LinX loops.

So once again, prove your overclocks are stable with P45. If you're so confident, there should be nothing stopping you from running LinX like I ask. Right now I think you guys arn't actually stable.

We dont have anything to prove that hasn't been prooved mate maybe you need to read some pages back...but its a known fact that giga UD3P's kick ass on quads...try to get yours at 500x9 stable at least and show the proof...either way show your screen at 520x9 that would be nice

[Zucker2k]

If you guys are so stable on gigabyte/p45 at these speeds, prove it by running 500 LinX loops, is what I'm saying. Because right now, I can show what you are all showing. My computer will run for 10 seconds and complete a superpi. It will also run for 8+ hours of prime95 at 520fsb, but it crashes after 30 or so LinX loops.

So once again, prove your overclocks are stable with P45. If you're so confident, there should be nothing stopping you from running LinX like I ask. Right now I think you guys arn't actually stable.

We dont have anything to prove that hasn't been prooved mate maybe you need to read some pages back...but its a known fact that giga UD3P's kick ass on quads...try to get yours at 500x9 stable at least and show the proof...either way show your screen at 520x9 that would be nice

You beat me to it Sergio. The guy is only stable at 457 and he wants to see what?

[andressergio]

hi andressergio. I see u have Gskill PI 1200 2x2GB ram. I got them too and I wonder what is the max voltage to run on them whitout active cooling and what max freguencies/ latencies can i expect from them?

you can expect a lot of them mate, i use them at 2,2 - 2,26 daily and benched them many times at 2,4+ CL5 you may want to read some pages back also what we've been posting with OC nub

Cheers mate

[zoson]

We dont have anything to prove that hasn't been prooved mate maybe you need to read some pages back...but its a known fact that giga UD3P's kick ass on quads...try to get yours at 500x9 stable at least and show the proof...either way show your screen at 520x9 that would be nice

I didn't say I was using the 9x multiplier, and in fact my screenshots on the last page SHOW I'm using the 8x multiplier.

Like I said, I can easily boot up to even 560FSB, but it is not stable. Taking a screenshot of superpi doesn't prove anything... 10 seconds of stress is nothing... So in reality, you do still have a LOT to prove.

[zoson]

You beat me to it Sergio. The guy is only stable at 457 and he wants to see what?

Yet you are the guys who are claiming to be stable at 550+ FSB and show no proof. Once again, 10 seconds of superpi stress is nothing. A screenshot of superpi or prime does not prove stability any more. Period.

And as you can see with my stability testing, I was at minimum 200 Linx loops stable at 550fsb on my e8600. Which is something NONE of you can say about your overclocks.

I'll boot up at 560 and do a superpi for you tonight. But honestly I think you guys all just call it a day and say 'stable' way too easily.

[andressergio]

I didn't say I was using the 9x multiplier, and in fact my screenshots on the last page SHOW I'm using the 8x multiplier.

Like I said, I can easily boot up to even 560FSB, but it is not stable. Taking a screenshot of superpi doesn't prove anything... 10 seconds of stress is nothing... So in reality, you do still have a LOT to prove.

mate we have nothing to prove to ya we just did it months ago

[andressergio]

You beat me to it Sergio. The guy is only stable at 457 and he wants to see what?

yeah ! i have a short tale

i was once upon a time a poor Asus user that couldn't get over that on FSB....

gigabyte
the end.

[zoson]

mate we have nothing to prove to ya we just did it months ago

But you didn't.
Saying you did something, when you in fact did not, makes you look like a poser.

If you did, repost the images. Why are you so scared to prove me wrong?

I've read all 112 pages of this thread. I had read them all before I posted.

[OC Nub]

Yet you are the guys who are claiming to be stable at 550+ FSB and show no proof. Once again, 10 seconds of superpi stress is nothing. A screenshot of superpi or prime does not prove stability any more. Period.

And as you can see with my stability testing, I was at minimum 200 Linx loops stable at 550fsb on my e8600. Which is something NONE of you can say about your overclocks.

I'll boot up at 560 and do a superpi for you tonight. But honestly I think you guys all just call it a day and say 'stable' way too easily.

All done on Gigabyte p45's, 2 different p45 giga boards,2 different c2q cpu's. a q9550 and a q9650.
If you take the time to read the threads you will see the stable oc's posted by many.

[Linuxfan]

yeah ! i have a short tale
i was once upon a time a poor Asus user that couldn't get over that on FSB....
gigabyte
the end.

Great story!
I consider Prime95 a fine way to test your CPU stability...
If your overclocks are stable enough for your uses, why do you need more?

[andressergio]

But you didn't.
Saying you did something, when you in fact did not, makes you look like a poser.

If you did, repost the images. Why are you so scared to prove me wrong?

I've read all 112 pages of this thread. I had read them all before I posted.

you should be of this Q9650 @ 4.115 9x457 457 is soooo maraca

you enter on a thread with 19 post attacking ppl i think if you run your asus your beloved 500 linx loops some mosfet will blow up and rip you an eye

[OC Nub]

But you didn't.
Saying you did something, when you in fact did not, makes you look like a poser.

If you did, repost the images. Why are you so scared to prove me wrong?

I've read all 112 pages of this thread. I had read them all before I posted.

Sergio has proven STABLE overclocks far beyond what your nvidia board will ever do using a 45nm quad. You come to this thread being an ass and really deserve no response.

When you top these oc's on your nvidia chipset board using a c2quad please come back and post your results. Until then learn to show respect to others in this great thread.

[andressergio]

Sergio has proven STABLE overclocks far beyond what your nvidia board will ever do using a 45nm quad. You come to this thread being an ass and really deserve no response.

When you top these oc's on your nvidia chipset board using a c2quad please come back and post your results. Until then learn to show respect to others in this great thread.

+1 for that

[Hoss331]

+2