The E0 Q9550 SLB9V thread (finally!)

[mngdew]

If I run my FSB any more than 444x8.5 there's no combination of VTT/GTL's AT ALL that will keep her running...

Have you tried setting CPU GTL to AUTO?

On my Rampage Formular, I only have 0.63x, 0.65x and 0.67x on CPU GTL. Like your system, I couldn't get my system stable, and getting the exact same BSOD message. Right be for I called it a quit, I just went, "what the heck, it wouldn't hurt anything at all", and set the CPU GTL to AUTO. Voila!

[5010]

Have you tried setting CPU GTL to AUTO?

On my Rampage Formular, I only have 0.63x, 0.65x and 0.67x on CPU GTL. Like your system, I couldn't get my system stable, and getting the exact same BSOD message. Right be for I called it a quit, I just went, "what the heck, it wouldn't hurt anything at all", and set the CPU GTL to AUTO. Voila!

Seems so silly but it worked! LOL

I had to shift one of my four lanes a tiny smidgen but voila! Thanks for suggesting auto dude!

[angelreaper]

My new q9550:

load:

idle:

[Ravynmagi]

First off, I'd like to shove a hot poker into the eyes of ya'll that got 4GHz with low voltage. Yes, I'm maliciously envious.

Anyway, I just got my second Q9550 since the first was so horrible (it had troubles at stock speeds). My second Q9550 seems to be doing much better. But it seems 4GHz is not gonna happen for me.

I'm at 3.8GHz now and Prime 95 ran for about 10 hours with 1.36 vcore and 1.20 vtt except for CPU3 which failed after 6 hours. Both GTLREFs are at 63%. Should I raise the second GTLREF up to 65%?

[Guacamole]

First off, I'd like to shove a hot poker into the eyes of ya'll that got 4GHz with low voltage. Yes, I'm maliciously envious.

Anyway, I just got my second Q9550 since the first was so horrible (it had troubles at stock speeds). My second Q9550 seems to be doing much better. But it seems 4GHz is not gonna happen for me.

I'm at 3.8GHz now and Prime 95 ran for about 10 hours with 1.36 vcore and 1.20 vtt except for CPU3 which failed after 6 hours. Both GTLREFs are at 63%. Should I raise the second GTLREF up to 65%?

Not the best board for OC'ing a Quad your vdroop is soooo bad I can't imagine you being able to get 4Ghz with that board and this chip unless you run it at about 1.5 and allow for the vdroop to take you to 1.4 or even worse 1.37 give or take.. Your problem is the board not the chip...

[VRSpy]

First off, I'd like to shove a hot poker into the eyes of ya'll that got 4GHz with low voltage. Yes, I'm maliciously envious.

Anyway, I just got my second Q9550 since the first was so horrible (it had troubles at stock speeds). My second Q9550 seems to be doing much better. But it seems 4GHz is not gonna happen for me.

I'm at 3.8GHz now and Prime 95 ran for about 10 hours with 1.36 vcore and 1.20 vtt except for CPU3 which failed after 6 hours. Both GTLREFs are at 63%. Should I raise the second GTLREF up to 65%?

Vtt 1.35v and try.

[Ravynmagi]

Not the best board for OC'ing a Quad your vdroop is soooo bad I can't imagine you being able to get 4Ghz with that board and this chip unless you run it at about 1.5 and allow for the vdroop to take you to 1.4 or even worse 1.37 give or take.. Your problem is the board not the chip...

I had been wondering about that and asking around if my P35 board would limit my overclocking with the Q9550. I almost picked up a P45 board, but I got the impression it wouldn't make a difference.

I have a vdroop of about 0.06 to 0.07. So my 1.36 in the BIOS drops to about 1.30 with occasional dips to 1.29 under Prime95 Small FFTs.

Is the vdroop on the P45 boards less?

[golfjam]

I had been wondering about that and asking around if my P35 board would limit my overclocking with the Q9550. I almost picked up a P45 board, but I got the impression it wouldn't make a difference.

I have a vdroop of about 0.06 to 0.07. So my 1.36 in the BIOS drops to about 1.30 with occasional dips to 1.29 under Prime95 Small FFTs.

Is the vdroop on the P45 boards less?

Nope but Asus And Gigabyte have "Loadline Calibration" option on bios that minimizes vdroop.

[zlojack]

This CPU has piqued my interest...I was thinking about the Q9650, but if these are getting up to 4-4.1 GHz, then I can put the extra $300 to good use elsewhere.

[angelreaper]

This CPU has piqued my interest...I was thinking about the Q9650, but if these are getting up to 4-4.1 GHz, then I can put the extra $300 to good use elsewhere.

No problem 4GHz with Q9550:

[IlludiumQ36]

Also as of 10/08 Newegg.com has started shipping E0's. Everyone that ordered after 10/08 so far that I've talked to has a E0.. It's probably not a For sure thing with Newegg, as they ship from all over. but They have moved plenty. as the Q9550 seems like the New Q6600.

Good Luck Guys!

I wish that were the case. I ordered 2 9550's last Friday and got 1-E0 and 1-C1. The C1 blows. In identically configured systems, it's 12-16° hotter than the E0 at load and I can't get it higher than 3.5 due to heat more so than stability. I've lapped and reseated just to be sure - no change in temps.

My new q9550:

load:

idle:

That is sweet. Can you share your settings? I have the same chip and motherboard in my office system and I'm coming up a bit short(can't prime past 440FSB).

[vws]

I wish that were the case. I ordered 2 9550's last Friday and got 1-E0 and 1-C1. The C1 blows. In identically configured systems, it's 12-16° hotter than the E0 at load and I can't get it higher than 3.5 due to heat more so than stability. I've lapped and reseated just to be sure - no change in temps.

Can you tell us the VID values of both 9550's you got? To correctly read the VID values, you can use CoreTemp.

[angelreaper]

That is sweet. Can you share your settings? I have the same chip and motherboard in my office system and I'm coming up a bit short(can't prime past 440FSB).

Here man:
CPU Voltage : 1.2750-1.2870
CPU GTL Voltage Reference (0/2) : Auto
CPU GTL Voltage Reference (1/3) : Auto
CPU PLL Voltage : 1.52
FSB Termination Voltage : 1.24
DRAM Voltage : 1.98
NB Voltage : 1.36
NB GTL Reference : Auto
SBridge Voltage : 1.10
PCIE SATA Voltage : Auto

Load Line Calibration : Enabled
CPU Spread Spectrum : Disabled
PCIE Spread Spectrum : Disabled
CPU Clock Skew : Normal
NB Clock Skew : Normal
CPU Margin Enhancement : Optimized

And little more hoot:

[IlludiumQ36]

Thanks, angelreaper. You've given me a push in the right direction.

Can you tell us the VID values of both 9550's you got? To correctly read the VID values, you can use CoreTemp.

I have CoreTemp installed on both machines. I'll pull that information when I return to work Monday and post it here.

[vws]

Thanks IlludiumQ36!

[ChaosAD]

Very nice clock you have there angelreaper, can u post some everest runs? Illudium do some tests and post some pics of the results with your MFII board if possible. Im really interested in some results

[golfjam]

My contribution. Batch # L834B475 Pack date 10/15/08.

[carepolice]

My contribution. Batch # L834B475 Pack date 10/15/08.

Umm did 3Dmark end up on that high temps what realtemp shows in "max temp" section?

Its pretty high for... duh atleast for me

Still

Very nice OC there

[golfjam]

Umm did 3Dmark end up on that high temps what realtemp shows in "max temp" section?

Its pretty high for... duh atleast for me

Still

Very nice OC there

Thanks. Those temps are after OCCT and benchies. It gets hotter than that running prime but max is 70~72 which is ok for air w/ 1800rpm fan running at 80%.

[IlludiumQ36]

Thanks IlludiumQ36!

CoreTemp is reading VID for BOTH chips at 1.200v. I would have thought there would be a minor difference.

Chip specs:
Stepping C1
S-spec - SLAWQ
Batch # - L825B512
MM# - 897363

Stepping E0
S-spec - SLB8V
Batch # - L827B287
MM # 899133

I have both machines identically configured in hardware and OC settings(currently set at 3.6Ghz), both 6 hours prime stable. E0 idles @ 29°C , C1 @ 44°C . That temp difference is consistent at load too.

Illudium do some tests and post some pics of the results with your MFII board if possible. Im really interested in some results

Do you mean test the 9550 on the MFII? Sorry, man. My MFII is retired from testing as it is now my home 24/7 rig. If you are wanting some settings and results on that setup(E8500/MFII), I'll do my best to find some system down time and run the benchmarks/tests.

[5010]

I ditched my E0 it was so brutal... 1.36VCORE loaded for 3.78GHz after a solid week of tweaking!

Grabbed a C1 with a 1.1325VID and its running 3.83 on 1.28Vcore after a couple of hours she's already 10x IBT stable

Damned E0 piece of trash that chip was

[vws]

@ IlludiumQ36: Great piece of info - thanks for sharing!

@ 5010: That C1's VID is outstanding. What's the VID for the E0, do you still remember?

[Peter Burg]

Here's a quite bad Q9550 .. but it's stable



Tomorrow my UT P45 arrives. Let's see how it is going to perform

[5010]

@ 5010: That C1's VID is outstanding. What's the VID for the E0, do you still remember?

was 1.25VID

While this C1 is a bit hotter than the E0 it uses so much less voltage..

[carepolice]

was 1.25VID

While this C1 is a bit hotter than the E0 it uses so much less voltage..

Umm.. even your old chip got higher VID that doesnt make it any tyer than any lower VID ones.

example:
1.15VID you clock it so hard that you need 1.35 VCORE so there is a 0.2V increase, And it runs really hot ofc

1.25VID You clock it and use 1.45 VCORE so there is a 0.2V increase, it runs same temps as 1.15VID one on 1.35 VCORE.

simple?

This is why i keep intels specified volt-range pretty ed up. It "limits" higher VID oc's. In my situation i keep 1.4V "max" for standard use 0.15 increase...