Q9550 & UD3R, Noob looking for stable settings,possible overclock later on

[stavagoliam]

Re: Q9550 & UD3R, Noob looking for stable settings,possible overclock later on

It was an operator's fault indeed, quite frankly I'm an idiot! I was trying to boot from the LS120 , which I don't have. I got confused/misled by the device manager's properties for my LG-DVD writer( HL-DT-ST DVD-RAM GH22LS30 ) and also the problem I had with the WD Lifeguard Tools boot cd I had earlier, i.e. "IDE CD-ROM device driver NOT installed". I'm really sorry for wasting all your time , I really need to go get myself right now!!!

[profJim]

Re: Q9550 & UD3R, Noob looking for stable settings,possible overclock later on

it happens to the best of us!!

[stavagoliam]

Re: Q9550 & UD3R, Noob looking for stable settings,possible overclock later on

Just a qwik extra on the temps,in the I wouldnt worry about variations too much vein.
My cpu has allways read that way on all utilities since I got it.
The first cpu hs i used was one that was on my D820 (it looked better than the Q8200 one),
I had that on and off the cpu a lot.I was tring "pad mods " on the processor different paste application methods n all sorts of weirdness.

There are posts about quad temp core variations if you check it out,on google etc.
People who are worried about it,but its fairly common.
Some get all 4 cores reading the same ,some get variations.
There are theories about "stuck sensors" etc and recalibration on utilities like realtemp.

re: temperatures

Some TIM (thermal interface material) takes longer to fully set up and reach optimal performance. Arctic Silever5 and others require quite a few heating/cooling cycles to fully cure and reach optimal efficiency. Running your system continuously slows this process. It's possible that you applied too much or too little TIM. At low loads, your cpu is probably using fewer cores. At full load, all cores are probably being used and the temperature variation between cores would be lower. As long as the temps are reasonable, don't worry.

re: cdrom booting

Both my systems use IDE cdroms and boot fine. Check the instructions and verify that your cdrom doesn't require a specific jumper setup.

Ive never had my cooler off the cpu since I got the cooler,3 months now(good job too cos itd be a pain to remove).
If Id never run the cpu before that Id be worried it wasnt seated properly or something,as it is tho..Im not really worried because of previous testing.
Core 0,or Core 1 depending how you look at it is the only one that reads hotter than the rest and I think of it as the "safety" valve.
I get a 10 degree max variation between the first and 4th core at high temps,had first core at 71 once core 4 was 61.
At standard (unclocked) cpu under load my core 1 was 57 core 4 was 47 ..that kind of thing.

Temperatures seem to have settled down quite a bit now( almost 10 degrees!! ). I try to test the new system as I have it on for about 7-8 hours a day, stress testing and such, then I shut it off for the rest of the day. It makes a lot of sense what Jim said before:
"Arctic Silever5 and others require quite a few heating/cooling cycles to fully cure and reach optimal efficiency. Running your system continuously slows this process."

After 10 hrs of prime95 I've got:

min: core0: 33, core1: 40, core2: 33, core3: 32;
max: core0: 47, core1: 47, core2: 45, core3: 43

After 1 hour of OCCT's CPU:OCCT temperature readings were the same as what 10 hrs of prime gave me.
However I got the highest temp readings on all 4 cores when I ran 1 hour of OCCT's CPU:LINPACK:

max: core0: 49, core1: 49, core2: 47, core3: 47

That's with Vcore of 1.16-1.17V.

[profJim]

Re: Q9550 & UD3R, Noob looking for stable settings,possible overclock later on

As I recall, it can take up to 200 heating/cooling cycles for thermal interface material to reach its full heat transfer capabilities. Some T.I.M.s will "cure" with fewer heat/cool cycles.

[stavagoliam]

Re: Q9550 & UD3R, Noob looking for stable settings,possible overclock later on

I see, I used the MX-2 compound which is identical to the Arctic Silever 5( just got the better reviews ). I'm glad I did! Its been less then a week since I applied it and I can already see 8-9 degrees temp drop at full load( no change @ idle tho). After all the stress testing I found that there is a stuck core sensor on my q9550. When idling it's value won't change - it stays at 40. When there's any load it pops up. I ran the RealTempSensorTest and it gave me "0" value on that same core( the others shoed 5, 3 &1 ). Well, I'll just ignore it for now and see if it will cause any problems when I try to overclock this puppy.