Overclock.net banner

4GHz Q9550 E0 - 90% IntelBurnTest

2542 Views 23 Replies 14 Participants Last post by  opt33
I've finally got my Q9550 E0 up to 4GHz passing 9/10 tests in IntelBurnTest.
Here are my settings:

471x8.5 = 4003GHz
CPU core: 1.34v
CPU PLL: 1.56v
FSB Term: 1.3v
DRAM: 2.1v
NB: 1.36v
SB: 1.1v

This peaks at 55-56c in IntelBurnTest which I am very happy about as other tests are usually 7-8c lower than this. Although I would like to pass 10 tests. Would anyone have any suggestions about tweaking voltages to get a more stable clock?
Heres a screen:


Im also weighing up going back to 3.9GHz which I can get fully stable in IntelBurnTest at 1.31v.
See less See more
1 - 20 of 24 Posts
agree'd
Really? From what I've read alot of people have said that IntelBurnTest pushes your system much harder than Prime95? What would be the differences?
in real life situations your CPU will never face what intelburn test puts it through.

prime95 for long periods of time is much more realistic.
(like folding would be)
2
Quote:

Originally Posted by paamlogs View Post
Really? From what I've read alot of people have said that IntelBurnTest pushes your system much harder than Prime95? What would be the differences?
I dont think ive seen anything better than Prime.. Even Skulltrails is crippled and falls down on all 8 cores when stressed
See less See more
So does that mean I may already be stable and I could have pushed it more than I need to? Because that would be good
See less See more
Intel burntest is better, when searching for 100% stability, only intel burntest can provide those results. Do know that Intel burntest burns about 7-15C hotter than prime95/orthos. 10-20 minutes on intel burntest is better than 12-24 hours of prime95, very handy when you are overclocking.

once again, some people may settle for "good enough" but to insure 100% stability, intel burntest is the way to go.
2
Quote:

Originally Posted by dennisjai View Post
Intel burntest is better, when searching for 100% stability, only intel burntest can provide those results. Do know that Intel burntest burns about 7-15C hotter than prime95/orthos. 10-20 minutes on intel burntest is better than 12-24 hours of prime95, very handy when you are overclocking.

once again, some people may settle for "good enough" but to insure 100% stability, intel burntest is the way to go.
propaganda has been "burned" into the soul of this one.



there is not one application out there that will stress your processor as much as intelburn test.

there is only ONE application in the world that stresses harder than small fft on prime95 or orthos. and that's intelburn test.
See less See more
Regardless of which program you use you still crashed.

So, up the NB and FSB voltages one to two notches and see if that stabilizes you. If it gets a little bit better, try upping them one more notch. If it is 99% stable after that you may want to give the vCore one more notch as well.
2
Quote:

Originally Posted by rpm666 View Post
Regardless of which program you use you still crashed.

So, up the NB and FSB voltages one to two notches and see if that stabilizes you. If it gets a little bit better, try upping them one more notch. If it is 99% stable after that you may want to give the vCore one more notch as well.
very true...
lol

there's a little bit of breathing room on both the NB and vcore.

remember not to go over 1.36v with a 45nm chip for vcore.
See less See more
Cool thanks guys
How far to you think I could go on the FSB and NB because I want to keep the Vcore down as much as I can?
See less See more
It's probably going to be that you need a bit more CPU Vcore, rather than NB voltage. But you may as well try increasing the NB voltage first, since you have somewhat more headroom there right now.
I don't trust burning the **** outa my new CPU for intel burntest is worth while, or even needed. Run Prime95 for 12 or 24 hours.. and be good.

Burntest is to much for me and I think it's overdone! and I've heard some horror stories of a few peeps CPU's always running slightly hotter after a 12 hour Intel Burntest session. Not sure though
2
Quote:

Originally Posted by xguntherc View Post
I don't trust burning the **** outa my new CPU for intel burntest is worth while, or even needed. Run Prime95 for 12 or 24 hours.. and be good.

Burntest is to much for me and I think it's overdone! and I've heard some horror stories of a few peeps CPU's always running slightly hotter after a 12 hour Intel Burntest session. Not sure though

Haha true but I am pretty happy with a peak of only 56c in BurnTest
All other benches I have tried fall right under that in terms of temperature. Im gonna try Pime95 small FFTs now and I might be able to drop my Vcore a bit to remain stable.
See less See more
Quote:


Originally Posted by xguntherc
View Post

I don't trust burning the **** outa my new CPU for intel burntest is worth while, or even needed. Run Prime95 for 12 or 24 hours.. and be good.

Burntest is to much for me and I think it's overdone! and I've heard some horror stories of a few peeps CPU's always running slightly hotter after a 12 hour Intel Burntest session. Not sure though

Oh your CPU went through worse before you ever got it, at Intel.
See less See more
Quote:


Originally Posted by xguntherc
View Post

I don't trust burning the **** outa my new CPU for intel burntest is worth while, or even needed. Run Prime95 for 12 or 24 hours.. and be good.

Burntest is to much for me and I think it's overdone! and I've heard some horror stories of a few peeps CPU's always running slightly hotter after a 12 hour Intel Burntest session. Not sure though

I disagree first of all, every time I try an OC I don't want to leave my PC on for a day to see if it is stable. Only to come back and find it is not, then have to leave it on for another day. IBT is just as good if not better than orthos/prime 95 and is much more convenient.
See less See more
Cool. I'm really debating on selling again, and picking up a 9550 or 9650.

Subscribed to see further results.
I demand absolute stability, even in harsh conditions, and I won't use an OC that will not pass at least 100 runs of IBT.

Even though the chances of you finding a real program that provides anywhere near the stress of IBT, that does not mean that you should consider a system that fails IBT stable.

The exact same factors that cause instability in IBT can, and probably will, cause computational errors in much less stressful applications, eventually. It may take an exponentially greater amount of time, but unstable is unstable.

No single test is going to give a difinitive answer on stability and if any stability test fails, something is wrong.
I was doing a quick 5 passes of Intel Burn Test to see if the CPU was (relatively) stable, and that saved me a day or so (each reboot) getting the voltages dialed in. After it passed I ran folding 24x7. (Still do in fact.)
1 - 20 of 24 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top