Q9450 OC / Temps / Settings

[lowtcell]

Well, LLC caused a lot of problems for people on the early Rampage boards (Read the early pages of the Rampage section) and i know this was/is a issue for more Asus boards.
24/7 Prime stable and random crashes during surfing or idle, problems where solved by disabling LLC and pencil mods.

I already have the board pencil molded. I did that back when I was running the E6750. When I switched out to a Q6600, the vdroop returned and I have been using LLC. With LLC disabled, I get .008 vdroop under load. Should I redo the mod?

[carepolice]

Whatkind of temps are you guys hitting on 1.35->1.4! (45nm ofc) My C1 1.40V (Just test) and intelburner burns my chip up to 70c, pretty nice IMO.

Oh well idle if it matters ~34c

[ewitte]

3.8Ghz @ 1.35v on q9550

AC set to 80F and fan on normal speed
~60C on all cores with Prime

AC set to 76F and fan cranked up a little
~56C on all cores with Prime

[Slovnaft]

Whatkind of temps are you guys hitting on 1.35->1.4! (45nm ofc) My C1 1.40V (Just test) and intelburner burns my chip up to 70c, pretty nice IMO.

Oh well idle if it matters ~34c

Really? ac or wc? cos that's pretty low IBT load temp for 1.4v...

[carepolice]

Air cooling, and yeah temps are from latest Realtemp 100tjmax ofc. Xigmatek is doing well.

[Slovnaft]

i don't know what to say.
that's incredible.
hard to believe though, especially if you idle at 34.
1.4v? you sure? i mean, at 1.3 my q9450 starts to hit 80 under IBT, and after that the temp/volt curve gets pretty steep.
I'm under tunitower and it's hard to believe there's that much difference.
i guess i'm glad i bought that xig for the family build ;]

[carepolice]

Well yes.... I got pretty good airblow on my antec. 1.36V 4h PRIME95 And temps were around 55-63. 1.4V Intelburner max temp was 72c

I am happy of these temps becouse my chip is 1.25000VID and needs pretty high voltages for high clocks

[shiznit93]

X38 is definitely a restricting chipset.
Don't get me wrong, E0 stepping cpus have definitely proven their worth, but I've been through 2 C1 Q9450s and they're both champion clockers.
So before you fry your Q9550 for an RMA, I'd consider picking up a nice Biostar Tpower I45 and really unlocking the 4+ghz potential of your cpu.
Because really, if my Q9450 can hit 4ghz, most Q9550s should be able to*

*p45 chipset required...

are you saying that p45 helps with FSB wall on 45nm quads? I got rid of my X3350 because I couldn't get it much past 3.3ghz on my x38 mobo.

[Slovnaft]

are you saying that p45 helps with FSB wall on 45nm quads? I got rid of my X3350 because I couldn't get it much past 3.3ghz on my x38 mobo.

uhh CHYAA. putting a proc on a p45 chipset is pretty much guaranteeing your clock isn't mobo-limited.
You probably shouldve gotten rid of the mobo instead of the proc...

[shiznit93]

uhh CHYAA. putting a proc on a p45 chipset is pretty much guaranteeing your clock isn't mobo-limited.
You probably shouldve gotten rid of the mobo instead of the proc...

well p45 wasn't out at the time and I got a sweet deal on the E8400.... but yea...

DOH!

[carepolice]

[Singh400]

3.6 Ghz is driving me nutty.

I do a P95 (8hrs +) and IBT x10. All passes everything is fine.

I reboot and do the exact same thing and IBT fails and so does P95 (randomly, rounding error).

Very reluctant to up the voltages, temps sit at 65-66c when IBt'ing. And I've tried to prochot switch at 71c...three times.

Any ideas?

Other than that, the system for usablity is stable, it's only P95 and IBT "tellin me" that my system is unstable.

[carepolice]

3.6 Ghz is driving me nutty.

I do a P95 (8hrs +) and IBT x10. All passes everything is fine.

I reboot and do the exact same thing and IBT fails and so does P95 (randomly, rounding error).

Very reluctant to up the voltages, temps sit at 65-66c when IBt'ing. And I've tried to prochot switch at 71c...three times.

Any ideas?

Other than that, the system for usablity is stable, it's only P95 and IBT "tellin me" that my system is unstable.

Thats just p45, If your machine works on your daily use why bother looking at stress test results?

[lowtcell]

I disabled LLC on Friday and left bios Vcore @ 1.3000, CPUz 1.280 idle, and now 1.272 load. So I get .008 vdroop, not too bad. Still no crashes so far, fingers crossed. (412x8.5 3.51ghz - RAM 495 5-5-5-15 1.98v)

[Slovnaft]

.008 vdroop aint bad at all!
you may be able to mod out that vdrop, but i wouldn't worry too soon.
that's a good clock for a quad, though it's a shame, with almost any recent board you could be getting 4ghz...
check out that spankin biostar board for a cheapish upgrade.

[Earzz]

I already have the board pencil molded. I did that back when I was running the E6750. When I switched out to a Q6600, the vdroop returned and I have been using LLC. With LLC disabled, I get .008 vdroop under load. Should I redo the mod?

Sorry for the late reply m8, yeah if the drop isnt to bad leave the pencil mod you can always do it again.
Hopefully im right with disabling LLC but it looks like i am , goodluck man.

[lowtcell]

Thanks guys. I'll keep with this for now and see how it goes. I don't want to spend any more money on this thing with the i7 so close. I do regret selling my Q6600. It was 3.8 stable and 3.96 max, but oh well, spilled milk, etc......

[Singh400]

That's it. Forget 3.6 Ghz stable. Core 3 throwing a wobble, bumped up voltages, Core 0 throwing a wobble, bumped up nb volt, Core 1 throwing a wobble within 7 mins.

I'm back to my 3.4 Ghz stable - Small FFTs and Blend both 12hrs+. And IBT stable.

Saying that, I will problem have another bash at it. I can't help myself, I can feel I'm close.

Regarding LLC, when I disable this, the system has a proper fit.

[Slovnaft]

aww cmon, you gotta be able to do 3.6 stable! try 1.6vnb and 1.3vcore? you positive your ram's stable?
I'd really hate to see a Q9450 crapping out like this! mine have shown me so much love!

[Singh400]

aww cmon, you gotta be able to do 3.6 stable! try 1.6vnb and 1.3vcore? you positive your ram's stable?
I'd really hate to see a Q9450 crapping out like this! mine have shown me so much love!

Yeah RAM is rock solid, I could nuke those sticks and they would still work!

3.6 is so close, yet so far, it's FRUSTRATING

[Earzz]

Yeah RAM is rock solid, I could nuke those sticks and they would still work!

3.6 is so close, yet so far, it's FRUSTRATING

Lol, yeah it can be really frustrating.
I cant go any higher myself no more, strangly enough i was able to boot 3.8 and 4Ghz when i just got this cpu and now thats not even possible anymore :S.
Never really tortured my cpu with high voltages but i did ad two HD's so i think that it causing the problem.
What speeds do you get out of your ram? I still didnt try to push mine so im a bit curious what they can do.

[Singh400]

I can boot to 3.8 fine, just not IBT or P95 stable at all. RAM OCs to 556Mhz easy peasy. Stable too. But I prefer to keep RAM at stock or less speeds while trying to obtain the magic 3.6

[Singh400]

I don't want to speak too soon, but I think I might have 3.6 Ghz stable, IBT passes fine. P95/Blending now.

[truehighroller]

Did you use my settings I posted for you? I reposted them in the P5Q thread for you the other day.

[Slovnaft]

Good on ya. Personally if i passed IBT i wouldn't bother with P95. guess it depends on how much RAM and how many passes though.

Mind sharing some volt info?