• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Need some help overclocking Q6600

S

spr1nt

n00b
Joined
Mar 23, 2008
Messages
19
My overclocked q6600 running at 3.2 GHz is stable after about 12 hrs of running prime95. Problem is after a couple hours my temps hit around ~65c according to speedfan and coretemp. Idle is about ~44c. My setup:

GIGABYTE GA-P35-DS3L
q6600 (G0 stepping)
xigmatek HDT-S1283 cooler w/AS5 applied
A-DATA 4GB DDR2 800
pny 8800GT
cooler master 330 case w/the side open

Heres my bios screen:

http://www.s208270823.onlinehome.us/oc/bios.jpg
http://www.s208270823.onlinehome.us/oc/bios2.jpg

If I drop my CPU Voltage anymore, I get errors in prime. Any suggestions on how I can lower my temps?
 
That is not high for load temps. I would not even start to be concerned until it was over 75C or better yet use the "options" in the menu and set it to read "delta to Tjuncion" and as long as the number is over 20C (it will count down as the cpu get hotter) there is little need for concern.

If it will do 3.2 at that low voltage you have tons of room for more. I would not hesitate to go up to 1.35V in small steps if needed (watching the thermal margin per above).

If it was me, I would be running 8 x 400 for 3.2GHz so that my memory speed was on spec and providing more bandwidth than your current settings. You will probably need to up the mch voltage (+.1 or +.2 and make sure there is good MCH/Northbridge cooling.) Then you can start minor bumps in FSB and see where you can get to without hamstringing your memory performance and actually gain from OCing your memory as well as the cpu. That board should do 400+ FSB if the MCH is cooled properly but sometimes people have issues. But it is worth a try and most certainly worth the trouble to redo the thermals on the MCH/add a fan if you are serious about an OC.

If you are happy with the 3.2 GHz as it stand now, eh, good enough, your temps are not out of line unless that is in an air conditioned room at 70F ambient. Then I would say the temps are a little warm but for a "normal" room around 80F not so bad. You really do not have a problem unless things get much worse when the side panel goes on.

edit.
After testing you will want to try "performance enhance" to FAST and retest, this lowers the Trd timing and can be a fairly decent memory subsytem improvment. If FAST works OK and does not introduce stability issues (the Orthos "Blend" test is a good one for that) what the hell, try "Turbo" . It just depends on your memory and the luck of the draw on the board. But see if you can do better than Normal while you are testing.

Take notes and change only one thing at a time.
 
I think you're ok actually. Given what you have, your options really are to either get a bigger case or go water. Anything else I doubt will make much of a difference.
 
If it will do 3.2 at that low voltage you have tons of room for more. I would not hesitate to go up to 1.35V in small steps if needed (watching the thermal margin per above).
Click to expand...

Sorry, what exactly do you mean by this? I thought the CPU voltage goal was to keep it as low as possible. Or do you mean I can OC it some more if I raise the voltage?
 
yep, more voltage usually means more overclocking without errors, but more heat and more cooling required to remove.

I think I'm running 1.4875V in bios, shows up as 1.41V in cpuid with vdroop.
Q6600 @ 3.6 (9x400) and 63C max under full load w Ultima 90 and a 120x38mm 69cfm fan.
 
BillParrish said:
That is not high for load temps. I would not even start to be concerned until it was over 75C or better yet use the "options" in the menu and set it to read "delta to Tjuncion" and as long as the number is over 20C (it will count down as the cpu get hotter) there is little need for concern.

If it will do 3.2 at that low voltage you have tons of room for more. I would not hesitate to go up to 1.35V in small steps if needed (watching the thermal margin per above).
Click to expand...

Completely agree with this. Using distance to TJMax is a lot more effective than trying to compare with other people getting temps from who knows where. The other day someone posted about their overclocked Q6600 running 20C idle and 45C at full load on air cooling, like that is normal, even though 20C is under ambient temp for most people... I think a lot of people are posting a bullshit temps from sheer ignorance or because they make up their own TJMax number.
 
spr1nt said:
Sorry, what exactly do you mean by this? I thought the CPU voltage goal was to keep it as low as possible. Or do you mean I can OC it some more if I raise the voltage?
Click to expand...

Both Logan321 and ocellaris are exactly correct.

There is a small community of people that enjoy seeing how low they can take CPU voltage and still maintain stock performance or a mild OC. Typically these machines have the primary "goal" of very quiet operation and/or very low heat output. Mainly for home theater and media center types of use where the machine is in the living room.

The vast majority of us have the goal of the highest possible performance. However that does not mean we ignore noise or heat, we just have to deal with them more aggressively, water cooling is a good example of that.

That said, by testing, once you reach whatever performance goal you are aiming for, it is a good idea to do additional testing and find the lowest voltage your CPU will remain stable while still maintaining the OC. One just does not bang in 1.45V of CPU voltage and see what happens. You test, raising the voltage in steps and seeing what additional CPU speed you can achieve. At some point it will become obvious that further increases in CPU voltage are either dangerous as you are reaching danger levels or the CPU will fail to respond to additional voltage with speed increases. But in general you do not limit the OC by forcing CPU voltage low. It is the standard practice, because it works, to increase the voltage significantly above stock to gain additional CPU speed.

So a lot depends of the definition of "goal". And it is not good to give the CPU any more voltage that it really needs to run at some frequency. But you should not hamstring the potential performance by restricting the CPU voltage to a low value unless you have a specific reason to need to do so. The last two years has shown us that Intel C2Ds respond wonderfully to extra voltage by giving clock speeds that are almost double the stock speed with no reliability or lifetime issues if done with care and reason. It is typcial to run 65nm cpus like the Q6600 at 1.4 Volts and some run them even higher. I personally based on intel specs and the experences of others set my personal limit at 1.4V but testing has shown that with my particular cpu, board, and memory anything over 1.375 provides only a tiny improvement and so I run 1.375 constantly. This was after weeks and weeks of careful testing and note taking. It was fast and easy to determine that around 1.375 was the sweet spot, but much longer to try everything else possbile with a slightly higher voltage and determine that increasing to 1.38 or 1.4 V really did not do much and being conservative and running the machine 18 hours a day the performance at 1.375 volts was the way to go.

I hoped this helped explain it. It just depends on what you are trying to do. Nothing wrong with a super low cpu voltage, just dont expect to set any cpu speed records.
 
You must log in or register to reply here.
Back
Top