Q9550 - 3.6Ghz Goal - Parts Arrive Soon

C

CedarScott

Limp Gawd
Joined
Nov 6, 2007
Messages
222
Done a few days of research thus far, trying to do my homework for this build.

My goal is to reach 3.6Ghz on my Q9550 processor on my P5Q-E motherboard and 8GB RAM.

Multiplier:8.5
FSB: 425 Mhz
Clock - 3612

The RAM: G.Skill 1066 PC2 8500 - 2 Kits - 8GB
This RAM has voltage spec of 2.0 - 2.1. I read someone has success DRAM Voltage: 2.06 with 8GB.

According to Intel's spec finder, the Q9550 voltage range is 0.85 - 1.3625V.

The question I have, what voltage do I need to START with to safely reach 425 FSB before I start going downward to hit cooler temps?

Will I need to worry about Northbridge / Southbridge voltages or is leaving these as auto safe for 3.6Ghz?

I plan to us AS5 as well. How long should I wait before attempting overclocks with the curing? A week? The AS5 guide says 200 hours... was thinking of running it 16 hours a day and shutting down every day for a week. Should be good enough?

Hardware in sig. Thanks!
 
Most people generally start with a low voltage and FSB and gradually raise them. However, if you'd like to see how high you can take it to begin with, I suggest you try with around 1.34v. That may or may not be enough, so be prepared to take it higher if need be. However, it's a good starting point and well within safe levels.

Whether or not you'll need to raise your NB voltage really depends on how your board and CPU work together, so there's no way to know in advance. Like the CPU voltage, raise as is necessary.

As for waiting for the AS5 to cure, I would say it isn't really necessary. It needs to cure in order to reach its best performance, however if your temps are okay right off the bat, there's no need to wait before you start to overclock. You could probably start tweaking around the minute you get the system running, although obviously it would be best to set up the system with your OS and programs first so you can properly test its stability.
 
Thanks for the reply.

I think my problem is, I'm not totally sure what a "low voltage" starting point is for this setup. I really would like to know what most people hit without touching voltages and just raising FSB.

It should be interesting as my first overclock attempt ever. I've read the guide on here twice and trying to grasp it all. I learn better hands on, so Tuesday I'll get to play out some plans.

Really, what I need is safe guidelines and that is what the guide provided.
 
The only way you'll really be able to tweak your chip is by playing around with it. Every CPU is different, so don't worry about what other people get and focus on what works best for you.
 
I do understand that. I needed to know what people were getting as a baseline to understand what is possible, what is "easy" and what is a good stable "medium" overclock with reliable cooling.

I don't want to top mine out as I want it to last 4 years or so. Therefore, 3.6 seemed to be a good medium overclock. However, the more I read, I might just go with 400FSB x 8.5 to start. Many hit that without having to change voltages.
 
If I wanted to do 400FSB x 8.5 Multi and get 3.4Ghz, would this be stock voltages with everything on auto or will I see problems?
 
Depends on how good a chip you end up with. There's a huge variation in terms of what you can get with different voltages when it comes to recent Intel chips. 3.4GHz could be doable on stock volts, or you might have to raise the voltage a bit to get it stable.
 
Can't believe there's that much variation between the same chips from Intel. Crazy. I'll gradually increase FSB from 333 to 400 and see what happens on Tuesday. Prime for about 1 hour right between minor bumps and 8 hours to test true stability?
 
CedarScott said:
Can't believe there's that much variation between the same chips from Intel. Crazy. I'll gradually increase FSB from 333 to 400 and see what happens on Tuesday. Prime for about 1 hour right between minor bumps and 8 hours to test true stability?
Click to expand...

Once you find the OC you want, if you want to test "true stability" then you would want to also max out your CPU and GPU to create more heat and system load (use Furmark) then run everything for at least 48 hours.

8 hours is not really much in the big picture. Lots of OCs can make it 2, 4, or 8 hours without a problem only to run into problems later. Thats why you see a lot of posts like "Ok one of my cores failed, but it was 17 hours into the burn test. Is this OK???" Well how OK that might be is completely up to the user. A lot of people run system that can not pass hours on end of Prime95, but it works fine for their games and the games do not crash so they consider it OK.

Whole point here is that you gotta decide on what your comfort level is. If 8 hours sounds OK to you, then go for it.
 
CedarScott said:
Can't believe there's that much variation between the same chips from Intel. Crazy. I'll gradually increase FSB from 333 to 400 and see what happens on Tuesday. Prime for about 1 hour right between minor bumps and 8 hours to test true stability?
Click to expand...

Actually, there's better than Prime. Try this: http://www.xtremesystems.org/forums/showthread.php?t=197835

It's a more user-friendly implementation of the same method Intel uses to stress-test their CPUs at the factory. Apparently, 8 minutes is about equivalent to 30-40 hours of Prime.
 
VN, I just joined that forum today too. Had a lot of info on my board there. Thanks for the link. Appreciated.
 
Zero82z said:
That's the idea :).
Click to expand...

I've been spreading this around a bit and its working great. Already found some failures in friend's PCs that they swore up and day "never crashed unless it was a game problem" :p Its definitely been weeding out problems faster than Prime95 testing from what I can tell so far.
 
yes my Q9450 was stable at 3.4 ghz in prime for over 12 hours. But only stable for an hour under IntelBurnnTest. Much faster than testing in prime but still use both!

I also found my PC to be burn test stable for 100 iterations with 4GB memory but I could get an error after 20 passes or so when using 8GB. you should be able to achive 400FSB, mine does that no problem its breaking the 400 FSB stable barrier I'm still working on.
 
You must log in or register to reply here.
Back
Top