Q9550 Questions

Associate
Joined
8 Aug 2007
Posts
820
Location
North West
Just upgraded my Q6600 to a Q9550 I got off the MM.

This chip runs WAY cooler (currently max temp 58degC full load @ 3.44MHz, my Q6600 @3.2GHz NON load was at 61degC :eek: )

Anyway my main question is what is the highest safe voltage for the Q9950? Intels site seems to say 1.3625V, but I've seen people with them running at 1.4000V and higher?

I'm currently at 3.44MHz but with the voltage at 1.3600MHz so I'm unsure wether to leave it at that or push higher. I know people have got these chips to 3.6-3.8 on air.

Any help/advice welcome! :D
 
Yeah, 1.368v is Intel's recommendation (under load, so consider voltage drop etc), but as long as the temperatures are ok, you should be able to go slightly over. My limit is 1.4v as temps get really warm!
 
Yeah, 1.368v is Intel's recommendation (under load, so consider voltage drop etc), but as long as the temperatures are ok, you should be able to go slightly over. My limit is 1.4v as temps get really warm!

OK thanks I'll up it a bit and see if I can't get 3.6, my temps arn't a problem ATM nice and cool :D
 
Are you sure its more vcore it needs at that speed? My q9550 seemed to love northbridge volts, but pushing the vcore too high too early killed stability immediately. It was strange going from prime stable to blue screen by increasing the voltage a notch, but Im sure that's what it was.

What sort of voltage is considered too high for 24/7 use on these, assuming temperatures are kept below 60 or so?
 
Are you sure its more vcore it needs at that speed? My q9550 seemed to love northbridge volts, but pushing the vcore too high too early killed stability immediately. It was strange going from prime stable to blue screen by increasing the voltage a notch, but Im sure that's what it was.

What sort of voltage is considered too high for 24/7 use on these, assuming temperatures are kept below 60 or so?

As said Intel recomend 1.3625 max but I'm at 1.3800, this chip seems to love volts, I've tried raising everything else but it just fails Prime until I up the volts even with NB etc at defauls :confused:
 
The point above which you start to get degradation is usually considered to be either 1.4V or 1.45V Vcore.
 
The point above which degradation happens is probably the important one. Any idea how it depends on temperature? I believe higher volts are acceptable when the processor is kept cold, which makes sense as long as electron migration is temperature dependent. I think it probably is, but don't fully trust my memory of quantum mechanics

edit: bad luck shadowake :(
 
if you set the GTL refs/votlages, and FSB term and cpll than you can drop that vcore off quite a bit I m only on 1.25volts for 3.83Ghz
 
if you set the GTL refs/votlages, and FSB term and cpll than you can drop that vcore off quite a bit I m only on 1.25volts for 3.83Ghz

Yeah going to try playing with some of the others voltages this morning and see if I cant get the vcore down and hopefully get the fsb a bit higher, I'd be happy with 3.6
 
Downloaded the latest BIOS for my mobo (IP35Pro) and managed to get the clock stable just over 3.7GHz, the half multiplyer really helped! Didn't think a BIOS update would make such a huge diference.
 
I believe higher volts are acceptable when the processor is kept cold, which makes sense as long as electron migration is temperature dependent. I think it probably is, but don't fully trust my memory of quantum mechanics

edit: bad luck shadowake :(

Temperatures have nothing to do with degradation, it's purely how many volts and for how long the processor is subjected to them.
 
As said, the q9550/q9650 e0 cpus do not need lots of vcore until past 4ghz. It's the other settings which counts to stability.

Not sure if q9550's are the same as q9650's but they should be and 99% of q9650's will run at 4Ghz at their stock VID (eg mine only needs 1.23v at 4Ghz) if you get the other settings right for your cpu/mobo.

Oh and for the records I am quite happy running mine at 1.41v 24/7 (now 4.6Ghz) as I don't beleive the whole degradation thing. I have even run mine at 1.58v (watercooled) just to get a benchmark at 4.95Ghz.
 
Last edited:
I'm currently at 3.44MHz but with the voltage at 1.3600MHz so I'm unsure wether to leave it at that or push higher.

I'm surprised that you need that much vcore for 3.4ghz. Most that I've seen, including mine, will do that on stock vcore of around 1.2v. I only need to bump to around 1.25v for 3.6ghz too. Sadly my cooling's not up to running it any faster than that 24/7.

What settings are you running now for 3.7ghz?
 
Last edited:
As said, the q9550/q9650 e0 cpus do not need lots of vcore until past 4ghz. It's the other settings which counts to stability.

Not sure if q9550's are the same as q9650's but they should be and 99% of q9650's will run at 4Ghz at their stock VID (eg mine only needs 1.23v at 4Ghz) if you get the other settings right for your cpu/mobo.

I wonder if I could tweak mine to be stable and run cooler at 3.6 or 3.8, with less vcore. Then I could have it at that 24/7.
 
I'll pull out my settings for my q9650 at 4ghz at 1.23v when i get home. That with the reduced multi should give you 3.77Ghz.

However, note that every cpu and mobo is different but it might give you some ideas.

Also check the massive q9550 overclocking thread at xstremsystems which has loads of templates for different mobos.

Also keep an eye of your nb temps on your abit ip35pro. I found that limited my overclock when I had a q6600 as the nb got too hot. The bolt mod for your board solves this as does putting a small fan on the nb cooler.
 
Mines IBT/OCCT/Linpack stable at 3.2GHz with stock voltages,
Its good to see what others are running as Im planning to see how far it will go =]
 
Oh and for the records I am quite happy running mine at 1.41v 24/7 (now 4.6Ghz) as I don't beleive the whole degradation thing. I have even run mine at 1.58v (watercooled) just to get a benchmark at 4.95Ghz.

Leave it at 1.58v and you'll soon believe in it :)
 
Back
Top Bottom