Q9450 w/ 3.8ghz stable

[Dopekitten]

Gotta love LinPack, tests the CPU so much faster. Although, im not sure what the standard test for stablity on it is, but it think i did pretty well. (I also ran Prime95 for like an hour, no problems)



Max temps were 72 C on the first core (you can see the rest in realtemp's max temp row). Note these are from linpack, which normally has temps about 10 C higher than prime95. (In prime i got max temps of 62 C for all the cores)

-I'm using a Q9450 C1 stepping 475x8 (who said E0 is better than C1?)
-Asus P5Q-E w/ 1406 modded Bios, board works like a dream
-Xigmatek S1283, and i have a fan from my old dell that goes up to 4k rpm, i soldered the old intel 4-pin connector to the wires (since dell's was proprierty) so i can modulated it.

Bios settings:

Vcore: 1.32 in bios, 1.296 actual idle, 1.312 actual load (i have LLC enabled)
VTT: 1.28
PLL: 1.50
GTL 0/2: .635
GTL 1/3: .67
vNB:1.28
vDRAM:2.1

Evreything else is stock (including all the advanced CPU settings, i don't know why people disable them)

Ill try lowering the voltage, or go for 4.0ghz
I'm sure i could get that, i'd jsut have to use Prime95 to test because i'd need more voltage and w/ linpack my temps would probably cause temperature protection to bump the mulit down to 6x.

If you didn't see already, this is all on air.

Anyway, does anybody else have a 3.8ghz+ Q9450? even if it's under water? Care to share your settings?

 

[eklipz330]

im impressed, good job

and just cause you hit 3.8 with a c1 does not make it better than e0...i've seen some ridiculous overclocks with the e0

 

[Dopekitten]

Show me a Q9450 at 3.8ghz+ E0 (or a Q9550 at 4.0ghz cause of the .5 extra multi)

I know you can get riduculous overclocks, but its like with E8600's, which, like the old E8400 (counterpart of my processor) already could get 4.5ish on air

 

[xnem3s1sx]

Yea, but I'm thinking that 72 degrees is a little hot, i wouldn't try to go any higher without water, or you may kill the CPU. BTW, you should include a pic of CPUID showing your overclock.

 

[Guest]

i have a C1 q9550 at 3.85 ghz rock stable...

and you can't call your system is stable until you do atleast 12 hours hours of prime 95 on all 3 modes... I can call mine stable b/c i've done that to mine...

many times systems fail at the 8 - 10 hour point... so testing for 1 hour doesn't mean sh*t

 

[Guest]

bullsh*t... I typed ^ that and then it didn't post it so i typed it again and then it double posted... how annoying

 

[eklipz330]

although i don't feel like searching for it, but i have seen cpu-z pics of a dude with his q9550 E0 at like 4.2, and the highest temp he hit was like 60 ish after 8 hours prime stable

i wonder how the q9650 does...havent seen too many oc's of that

 

[Dopekitten]

I know, 1 hour is crap
But i used linpack

http://www.hardforum.com/showthread.php?t=1341978

Take a look, lots of good info discussing it, you should be happy, this means you don't have to run prime95 for 24 hours anymore--next time read my post, i said that i ran linpack and that was what the screenshot was for.

Also, what were your bios settings? And run linpack, cause you can do 24 hour prime95, but still fail linpack

(also on a short note, if you don't want to read the threads, linpack, or intel burn in test is the test that intel uses to insure stability for its processors, and obviously it doesn't have time to run prime 95 for 24 hours =P)

Also, we don't see instable stock processors very often do we, so i'd say it works pretty well

@ekilpz, ok i belive you, you got me there then, but i still don't think the E0 is that much better than the C1

Also, as for the Q9650, maybe somebody will stick in an overclock here. I'll take a guess and say 4.4/4.6 is probably plausible for either stepping, considering the QX9650 could get 4.2 easily (most of them).

 

[aznguy0028]

was curious, whats your VID on your Q9450? can you please post. thanks

 

[Dopekitten]

Mine is 1.25, next time i post i screeny, ill have realtemp showing it.

 

[Onus]

"You hear a loud zzzzzzzzzzip!"

"What do we see"

"As you come over the hill, you see a great big e-dong waving aloft."
================
Nice job (assuming it holds up under testing).

 

[Dopekitten]

Yea, it does--if you guys want, i can run it under prime95 for however long you want, to prove that it is stable, with something you are familiar with.

Alright, well at 3.863 (i was going for 3.9ghz, but i did bad math ><) even at 1.345v in bios, fails intelburnintest

4.00ghz post's but doesn't boot into windows (at the same voltage). I'm to scared to go to 1.36

 

[Guest]

didn't know linpack is any good... sorry =P

anyway i have a vid of like 1.28... and i'm stable at 1.25 volts @ 3.85 ghz lol

i've got the GTL stuff at auto for now... temps aren't too bad... not going to really mess with the settings in the bios until i go water... hoping to get this thing past 4 ghz...

I already break 18 k with the q9550 and a 4870... but I just wanna get past the 4 ghz line

 

[Dopekitten]

Nice, i've gotten 17k w/ mine, but my 4870 is a bit of a wimp, i've overclocked it, but its kinda...well...eh...i dunno, not really above par

 

[aznguy0028]

Yea, it does--if you guys want, i can run it under prime95 for however long you want, to prove that it is stable, with something you are familiar with.

Alright, well at 3.863 (i was going for 3.9ghz, but i did bad math ><) even at 1.345v in bios, fails intelburnintest

4.00ghz post's but doesn't boot into windows (at the same voltage). I'm to scared to go to 1.36

first off, good job on the 3.8ghz mark. imo, i don't think you need to prove to me or anyone here that your comp is stable. if it is, more power to you. but i'm sure people would want to see some results on staple programs akin prime95.

i've read many Q9450 forums and i have not yet seen one hit the 4ghz mark stable. people can post and run cpu-z and stuff but no one i have seen got it to be stable under air.

also the voltages going up to 3.9-4.0 is ridiculously high for the 45nm's, above the 1.365 recommended range. if you can get it stable at 4.0ghz, i'm sure many many people would want to see that happen and your settings too.

and for the E0 stepping compared to the C1, in all the result's i've seen on OC-ing E0 chips, they go much higher than their C1 counterparts. ex. E8500 & Q9550

 

[Dopekitten]

Well im running Prime95 now, so ill post back later today then.

I think i might be able to get 3.9ghz stable with 1.365, but im not sure if that voltage is applying to max actual voltage or max voltage in BIOS. Otherwise, since i have almost no vdroop on this board w/ LLC, w/ 1.365 volts, would be "in" the range, but out when you looked at my actual vcore. This is why i'm afraid to do 1.365, or close to that.

1.34 was just to see if i could post at 4.0ghz with that

 

[V3NOM]

haha i stil dont see the point of any of those extreme stress tests if no matter wat u do it doesnt get any where near those temps every day... two virus program scans+folding@home+counter strike source = lower temps than prime/orthos and thats as far as i would usually go lol

 

[cd14]

Will someone show me a "E0" quad? Aren't E0's dual core E8400..E8600? The latest quad revisions (R0) are on the 8200, 9400 aren't they?

 

[Guest]

Well... perhaps a lot of voltage is required... but i have mine at 1.25 volts stable @ 3.85 ghz... So i don't know if 1.36 is really necessary for a quad to get to 4 ghz...

I mean I might just have a really good chip... I assume I have the C1 revision... I haven't actually checked... because at the time i bought it the E0's were just coming out.

 

[aznguy0028]

go here, has picture of E0 running in CPU-Z

http://www.ocxtreme.org/forumenus/showthread.php?t=3428

the first E0 stepping was introduced in august of this year on the Q9550.
the Q9550 has 2 steppings, the previous one was the C1 but now has the newer E0

the Q9650 is a native E0 processor, it has no other steppings at this moment

The E8400 also has a C0 stepping, but came out with the E0 recently as well. same thing with the E8500 variant, both C0 and E0 are available

the E8600 itself is a native E0 stepping only.

you are correct about the R0 on the Q9400 though, but that is only where they exist. the 9450-9650 lines are different.

 

[cd14]

go here, has picture of E0 running in CPU-Z

http://www.ocxtreme.org/forumenus/showthread.php?t=3428

the first E0 stepping was introduced in august of this year on the Q9550.
the Q9550 has 2 steppings, the previous one was the C1 but now has the newer E0

the Q9650 is a native E0 processor, it has no other steppings at this moment

The E8400 also has a C0 stepping, but came out with the E0 recently as well. same thing with the E8500 variant, both C0 and E0 are available

the E8600 itself is a native E0 stepping only.

you are correct about the R0 on the Q9400 though, but that is only where they exist. the 9450-9650 lines are different.

Thanks for the info!