My q9550 e0: 1.2125 VID, 3910Mhz@1.25 BIOS (1.232v CPU-Z)...FSB Wall @ ~470???

[jeeka]

Hey everyone,
I went to Fry's (San Marcos, CA in San Diego County) and convinced the "Cage Girl" that I needed her to get me an e0 q9550... I ended up with this:

q9550 (slb8v)
FPO/Batch #: L835B078
1.2125v VID

Right now, on my P5K Deluxe and at 460fsb (3910Mhz @ 1.25v BIOS, 1.232v in CPU-Z)...

my q6600 sat at 475fsb, but it seems that my machine is having issues while booting into Windows at anything above 470 with this q9550. It boots in and I eventually get a 0000009C BSOD.

So I upped the vCore to 1.35v and booted again. I know, jumping from 1.25v BIOS to 1.350v BIOS is overkill, but I wanted to remove the possibility that it's not vCore. Same result.

I know I have a really good chip compared to what others are able to reach with this stepping (awesome vCore and temps) but I might be FSB limited...

Can you take a look at my BIOS settings and let me know here I can adjust stuff, especially around the vFSB, GTL's and other voltage settings? I don't fully understand the relationships between all these Term Voltages, etc...

By the way, I haven't done any priming/stability stuff yet,but I have encoded an entire movie in Vegas 8.0 to iPhone Video (mp4), having the Cores between 80% and 100% for the entire time. (Load temps are in the mid 40's!!!) My load temps on the q9550 are LOWER THAN THE IDLE temps of my q6600!

Thanks!

[deachus]

I would try these settings. I run a Q9450 up to 3.8Ghz 475FSB with these settings no problem.

Vcore - 1.3v-1.33v
Vtt - 1.32V - 1.36v
PLL - 1.6v
CPU GTL - .65-.67

[Zucker2k]

1. Lower your pll to 1.50
2. Raise vtt to 1.40v


What memory are you running?

[SoulsCollective]

vTTs of 1.36 and 1.4V? Be very careful there, those are way past the recommended limits for 45nm chips.

[Zucker2k]

vTTs of 1.36 and 1.4V? Be very careful there, those are way past the recommended limits for 45nm chips.

And the RECOMMENDED clock is 2.83Ghz I own this chip, albeit a C1, and if you want to see a higher clock, you'll have to raise the vtt. It has already been proven high vtt has no ill effect, and I can attest to this. Since you're running a low gtl (x.063, best for this chip at high fsb) you'll need to run mid-range vtt. My 4.2Ghz is with x.067 x 1.44vtt @ 8.5x495 or x.063 x 1.70vtt. The latter is completely stable. Don't listen to what doomsday prophets are saying. Keep your eyes on your cpu pll however, above 1.58 - 1.60v, you'll be seeing instability.

[SoulsCollective]

And I own a Q9450, and would make sure that if I told someone to run a vTT way beyond what other users are blaming for causing CPU death that I mentioned the risks. I'm also not aware of any "proof" of no ill-effect, when Anandtech wrote an article specifically warning against vTT above 1.3V. My 3.6GHz is with a 1.24 vTT, so it's not always necessary.

[jeeka]

1. Lower your pll to 1.50
2. Raise vtt to 1.40v


What memory are you running?

RAM: 4GB (2x2GB) (5-5-5-15) 5:6@552Mhz (so far) 2.2v G.Skill F2-8500CLD5-4GBPK (In Sig)...

Any thoughts?

[NapalmV5]

whomever started all these v myths sure were very effective: strike fear and they listen

ive been running my 4.8ghz q9650 @ xtreme voltages for 2 months now.. no degradations.. no fatalities

[jeeka]

Hmmm ok, I'll tinker with a lower vTT first and report back... I really have a great chip I think and it would suck if it was FSB limited to ~470

[NapalmV5]

you got a good vid q9550.. 533fsb easy

[jeeka]

you got a good vid q9550.. 533fsb easy

Hah! Thanks Napalm. Let's hope I can get more out of this. Is it too early to start asking if the MB is the issue? It held up my q6600 at 475fsb and that was about it, 480fsb would freeze on boot.

Maybe a P5Q Deluxe (p45) might need to be in my Newegg cart soon?

BTW, my vtt/vfsb is at 1.3v now and no ill effects.....yet.

[Zucker2k]

And I own a Q9450, and would make sure that if I told someone to run a vTT way beyond what other users are blaming for causing CPU death that I mentioned the risks. I'm also not aware of any "proof" of no ill-effect, when Anandtech wrote an article specifically warning against vTT above 1.3V. My 3.6GHz is with a 1.24 vTT, so it's not always necessary.

I've run a Q9450 at 3.6Ghz with 1.20v vtt; it's simple, every chip is different! About Anandtech, they were specifically talking about A C0 QX9650; after that incident, there has been NO REPORTED 'deaths' whatsoever. On the other hand, Benchzowner has done an extensive test, over 45 days, testing chips at high vtt, and they did not degrade or die. Please keep the fear-mongering to a minimum.

[SoulsCollective]

Of course every chip is different. That's why you can't just tell someone to do X,Y and Z and it will work - just as you shouldn't tell someone to do something without mentioning that it may be dangerous. That's not fear-mongering, it's just being responsible. Some here are comfortable with running 1.9V through a CPU, great, but you can't tell someone to just go ahead and do it without mentioning the risks. Similarly, some here have reported chip death from vTT, some have not. If you're comfortable, great, but allow others to make an informed decision.

[Zucker2k]

RAM: 4GB (2x2GB) (5-5-5-15) 5:6@552Mhz (so far) 2.2v G.Skill F2-8500CLD5-4GBPK (In Sig)...

Any thoughts?

I have your ram, be sure you're able to ran them at those frequencies, G.Skills been binning rather tight. Those are definitely going to be your main impediment unless you resort to 1:1. Set TRFC to Auto for 52.

Your full timings should be:

5-5-5-3-52-6-3

8-3-5-4-6-4-6

14-5-1-6-6.

[Zucker2k]

Of course every chip is different. That's why you can't just tell someone to do X,Y and Z and it will work - just as you shouldn't tell someone to do something without mentioning that it may be dangerous. That's not fear-mongering, it's just being responsible. Some here are comfortable with running 1.9V through a CPU, great, but you can't tell someone to just go ahead and do it without mentioning the risks. Similarly, some here have reported chip death from vTT, some have not. If you're comfortable, great, but allow others to make an informed decision.

Chill out already; the person needed help, and so I pointed out what was holding him back, and you show up spinning speculation. In YOUR opinion, 1.4vtt is dangerous; as far as I'm concerned it's perfectly safe (I'm running 1.70v vtt); maybe you should warn him clocking his chip above 2.83Ghz is dangerous too. Obviously the guy knows what he's doing, I'm just pointing out where the bottleneck is. You're being very arrogant since you don't even have the chip in question. And your argument is based on SPECULATION!!!!!!!!!!!!!!

[SoulsCollective]

lol chill out? Fearmongering? Spinning speculation? Arrogant because I don't have a Q9550?

A respected review site has warned about their experiences running a 45nm quad at a vTT higher than 1.3V. There have been multiple posts in these forums about users reporting chip death. Intel has posted an official warning about vTT burning out chips. That's not speculation, and unless you're maintaining that there's some magical difference between a Q9450 and a Q9550, I'm just as entitled to comment as you are. We have a thread that is showing that some have found the exact opposite. That's great, but it doesn't mean that suddenly the issue goes away and everyone should pump 1.7 vTT through their chips. Will it allow them to clock higher? Of course. Is it dangerous? Quite possibly. Should people be allowed to make an informed decision? Of course. Did I say not to do it? No, I merely said that they should be careful.

[Zucker2k]

Intel also warns against overclocking, yet I didn't see you warn him against that? You'll rather pretend you know it all, by pointing out something that has been shown to be nothing but a false alarm. http://www.xtremesystems.org/forums/...d.php?t=198614 I believe this test, because it was purposely done (with insane volatges, mind you) to kill these chips and they survived intact with no ill effects). You can believe what you choose. By the way, that 3.6Ghz oc of yours is dangerous, you better revert to stock or you'll kill your chip

[SoulsCollective]

Learn to read before you attack others. I have never said that I know any better than you do. You call me arrogant, a know-it-all, why I don't know, when it is you who are ignoring contrary evidence. The thread you linked to, I already posted, affirming as you say that yes, yes some users are not having any issues. That does not make the comments by other posters and reviewers, saying that there is a problem, go away.

I advocate nothing except informed decisions. I have no agenda here, I couldn't care less if the OP kills his chip or not, although I'd love for them to be able to run a high vTT and get the results they want without any issues. All I want is that all possible issues be laid out, so people can make an informed decision. As you'll see if you search my post history, in another thread in which another poster told a different thread starter that anything over 1.1 vTT would kill their CPU, I was making your arguments, pointing out again the contrary evidence.

I see little point in continuing this discussion when you appear not to be even reading my posts. If you want to respond to me, I'd suggest you try via PM, as I won't be replying further in this thread.

[NapalmV5]

Hah! Thanks Napalm. Let's hope I can get more out of this. Is it too early to start asking if the MB is the issue? It held up my q6600 at 475fsb and that was about it, 480fsb would freeze on boot.

Maybe a P5Q Deluxe (p45) might need to be in my Newegg cart soon?

BTW, my vtt/vfsb is at 1.3v now and no ill effects.....yet.

heres a quick simulation on my ep45 dq6


500x9 needs less voltages.. higher fsb ~ higher volts

depending on the mobo.. your q9550 may need more or less

others have gotten 533fsb @ lower voltages than me

not forcing anyone into such voltages.. just showing you what volts it takes to get such clocks

indication only

wish you luck

[Zucker2k]

I came down hard on you because you're fond of propagating this falsehood. If vtt is so dangerous as you say, people like NapalmV5 who's pumped 1.8v+ vtt into his chip should have lost it a long time ago. Or is vtt a selective killer? I'm running 1.70 vtt with no ill effects either, at 4.2Ghz. I guess that's the difference between your oc and mine? Sorry, I take the 1.70v vtt

Finally, let me just say this: All overclocking is dangerous! There's no way around it. I didn't have to point that out to you though.


@ OP: Sorry we hijacked your thread. Believe me, you'll find more fsb if you up your vtt from 1.30v, there's no way around it with this chip, except higher gtl; but even that'll run out quickly. The vtt is there for a reason; the stock vtt is supposed to support 1333 fsb; you're way past that now. Goodluck.

Edit: Oh your NB at 1.7, is high enough, it should be able to support you up to around 495 / PL 9/10.

[Zucker2k]

heres a quick simulation on my ep45 dq6


500x9 needs less voltages.. higher fsb ~ higher volts

depending on the mobo.. your q9550 may need more or less

others have gotten 533fsb @ lower voltages than me

not forcing anyone into such voltages.. just showing you what volts it takes to get such clocks

indication only

wish you luck

Damn, Napalm, what ram is that? 2.56v?

[jeeka]

Thanks dude, I do know what I am doing, but I don't know ALL of what I am doing ...so that's why I call on my bros here....no need to fight guys, I understand you both are passionate, but we can disagree and let me burn my chip, if I so choose to take the risk.

I will try toying with these voltages/refs and ram timings and will report back. I am hoping to extract the most information as I can from this post, and also as a side effect, save someone else from posting a new thread if they can learn from this one.

Any parting words before I get really adventurous?

[NapalmV5]

Damn, Napalm, what ram is that? 2.56v?

yeh @ 534x9 i got to keep it @ 2.6v 2GBx2 mushkin 1066/gskill 1100

my p45 is an A2.. the A3s here on the forum.. a lot less vdimm @ same mem clocks

[AzN_plyR]

Nice chip, I got the same batch and everything but 1.875 VID also from Fry's Anaheim.

http://i58.photobucket.com/albums/g2...Quadstable.jpg

Fully stable primed/intelburn/occt stable.

Now I gotta get my ass up and try higher.

[andressergio]

Hey everyone,
I went to Fry's (San Marcos, CA in San Diego County) and convinced the "Cage Girl" that I needed her to get me an e0 q9550... I ended up with this:

q9550 (slb8v)
FPO/Batch #: L835B078
1.2125v VID

Right now, on my P5K Deluxe and at 460fsb (3910Mhz @ 1.25v BIOS, 1.232v in CPU-Z)...

hey jeeka nice vid...

I have a Q9650 L825B513 vid 1,25 testing on my Asus P5K deluxe bios 0902 also...yes Q9550 for this boards will limit for sure your max fsb as you onlu have 8,5multi...that's why i decided to buy this that has a x9 multi, of course there's a difference in price but i knew the limits of this board and have all the water cooler stuff i am lazy to change mobo right now, and my goal is 4000 or 4050

I also have a friend with a Q9550 L827B batch vid 1,25 on a rampage and we're both testing vcore, vtt, and nb to get as a reference...and i discovered that the P5K deluxe need just a notch of vcore more to get it stable...but he needs more vtt and nb than me, of course i don't have intermediate values to finest adjust my OC the rampage yes...

For now i'm primming smallfft's and hour on this settings

4005 (445x9) 2x2GB Mushkin Ascent 1066 2,05v 5-5-5-15 (trfc)= 55

static read = enabled
LLC = enabled
Transcaction Booster = Auto

vcore set on bios = 1,3125
PLL= 1,5
vtt (fsb) = 1,3
nb = 1,4
sb = 1,05
cpu ref 0,63
nb ref 0,67

temps LOAD 55/49/53/53 on water

Will post some screens later...

Cheers mate
Sergio