Q9550 OC not going as planned

g00mbs

n00b
Joined
Oct 29, 2008
Messages
19
Hey,

I've been reading two different forums to get my new rig up and running at a decent OC.

The to posts are http://www.hardforum.com/showthread.php?t=1198647 and http://www.evga.com/forums/tm.asp?m=499530&mpage=1&key=񹽊

My Bios Settings are:
BIOS Version: E179 SZ13 (I don’t know what P0 it would be)
**System **
CPU Freq, MHz: 2975 GHZ
FSB Reference Clock, MHz: 1400
CPU Multiplier: [8.5X]
PCle x16-1 & 16-2, MHz: [100]
PCle x16-3, MHz: [100]
SPPc->MCP Ref Clock, MHz: [200]
nForce SPP --> nForce MCP: [5X]
nForce SPP c-- nForce MCP: [5X]
**Spread Spectrum**
CPU Spread Spectrum: [Disabled]
HT Spread Spectrum: [Disabled]
PCle Spread Spectrurn(SPP): [Disabled]
PCle Spread Spectrurn(MCP): [Disabled]
SATA Spread Spectrum: [Disabled]
**FSB & Memorv Config**
FSB - Memory Clock Mode: [unlinked]
FSB - Memory Ratio:
FSB (QDR), MHz: [1400]
Actual FSB (QDR), MHz:1400
x MEM (DDR), MHz: 1600
Actual MEM (DDR), MHz: 1600
P1: [AUTO]
P2: [AUTO]
**Memory Timing Setting**
Memory Timing Setting: [Expert]
tCL (CAS Latency): 7
tRCD: 7
tRP: 7
tRAS: 18
Command Per Clock: [Auto(2T)]
** Advanced Memory Settings **
tRRD: [Auto]
tRC: [Auto]
tWR: [Auto]
tWTR: [Auto]
tFAW: [Auto]
tREF: [Auto]
**System Voltages**
CPU Core: [1.35 to 1.4]
CPU FSB: [1.4]
Memory: [1.9 to 1.925]
nForce SPP: [1.42 to 1.52]
nForce MCP: [1.5 to 1.65]
GTLVREF Lane 0: [Auto]
GTLVREF Lane 1: [Auto]
GTLVREF Lane 2: [Auto]
GTLVREF Lane 3: [Auto]
**CPU Configuration**
Limit CPUlD Maxval: Disabled
Intel Speedstep: Disabled
xPPM Mode: [Native]
CPU Thermal Control: [Disabled]
C1E Enhanced Halt State: [Disabled]
Execute Disable Bit: [Enabled and Disabled]
Virtualization Technology: [Disabled]
CPU Core 0: [Enabled]
CPU Core 1: [Enabled]
CPU Core 2: [Enabled]
CPU Core 3: [Enabled]

At a very minimal OC noted above in the bios settings, Core1 and Core2 are failing Prime95 large FFT immediately after starting the program. I have tried a range of voltages and am up to 12 different iteration of voltages sets with the same thing happening. Either those two cores fail right away or the rig crashes as it loads windows. Either way it isn't good.

My apartment is hot, I'm not gonna lie, my wife likes to keep it warm.

Originally I was trying to hit an OC of 400 MHz/1600 MHz, but all that would happen is crashes when loading windows.

Basically, I don't know what else to do cause going from 333/1333 to 350/1400 isn't really that big of a deal and I would probably just keep the CPU running at stock speed. It would be nice to OC, so any help would be appreciated.

~g00mbs
 
Ya, that's true. Basically I would like to get my FSB and MEM 1:1. I was thinking it could be possible with this cooler.

For FSB:Mem = 1333:1600, Mem @ 7-7-7-18 2T
CPU Vcore = 1.25v
Vtt = auto
SPP (NB) = 1.42v
MCP (SB) = 1.6v
MEM = 1.9v
CPU PLL = 1.5v

It seems to run stable. Considering that it usually hovers around ~80F in the room, the temps are:

Core 1 = 57 C
Core 2 = 50 C
Core 3 = 54 C
Core 4 = 54 C

But all in all I'm new to OCing.
 
to overclock a q9550, you have to bump up your vtt. auto, which is like 1.12v on your board is just not going to do it. i have mine set to 1.48v. also, you will probably need to tweak your gtl. and the 790i is really not the best board to overclock the 45nm quad. in fact, the only board that overclocks it well is the p45. so if you are serious about getting a decent overclock (3800 and up) give it some more vcore, a lot more vtt, and try a couple different gtl settings. its not like an e8400/8500 that all you do is raise vcore and set your fsb to 450. its takes a little finesse.
 
Don't bump your VTT too high. Anything above 1.4 is too high.

I have my Q9550@3.4Ghz@1.128vCore, 1.20 VTT
 
Don't bump your VTT too high. Anything above 1.4 is too high.

I have my Q9550@3.4Ghz@1.128vCore, 1.20 VTT

What are you using for cooling?

After posting on the EVGA forums I backed my Bios down to SZ10, but still no dice with 1600 FSB.
 
I use a Thermalright SI-128SE.

I can boot into 3.8Ghz easily at like 1.25 vcore, but temps go too high. Propbably due to crappy airflow in my case but I'll be upgrading to a ATCS 840 and liquid cooling soon to break the 4ghz mark :)
 
Oh I know I don't need WC. It's just I've been planning on building one for ages and I plan to push my vcore to probably 1.35 and see where I can go from there.

Temps being too high is at 1.25 vcore, if I run prime95, I can break 80 C. Don't even ask about IBT...
 
Thats odd...

When I'm running stock out of the box cpu speed 1333 with my memory unlinked at 1600 the following are my voltages set auto:

Vdroop = On
CPU Core = 1.25 V
CPU FSB = 1.1 V
SPP (North Bridge) = 1.42 V
MCP (South Bridge) = 1.5 V
CPU PLL = 1.5 V

For Mem I set the Voltage at 1.9 V to get the rated timings of 7-7-7-18 @ 1600.

Under the load of Prime95 running Large FFTs Everest reads the voltages as:

CPU Vcore = 1.19 V
SPP = 1.44 V
DIMM = 1.87 V

This is with two different Bios (SZ10 or SZ13) for the EVGA E179 790i FTW Mobo.

Temps read (SZ10):
Core 1: 54C
Core 2: 49C
Core 3: 52C
Core 4: 52C
SPP: 64C
MCP: 75C

Temps read (SZ13):
Core 1: 59C
Core 2: 54C
Core 3: 57C
Core 4: 57C
SPP: 62C
MCP: 69C


What I'm trying to get at here is that when I OC even a little bit I really have to push the CPU Core voltage up... in the range of 1.38 to 1.4 to even get the rig into Windows. The readings in Everest correspond to 1.36ish +- 0.1.
 
Your cores are looking a bit hot, is that on idle?

You really don't want to have auto settings. Manually pick out settings.
 
This is my Q9550

Q9550OC.png
 
Thats odd...

When I'm running stock out of the box cpu speed 1333 with my memory unlinked at 1600 the following are my voltages set auto:

Vdroop = On
CPU Core = 1.25 V
CPU FSB = 1.1 V
SPP (North Bridge) = 1.42 V
MCP (South Bridge) = 1.5 V
CPU PLL = 1.5 V

For Mem I set the Voltage at 1.9 V to get the rated timings of 7-7-7-18 @ 1600.

Under the load of Prime95 running Large FFTs Everest reads the voltages as:

CPU Vcore = 1.19 V
SPP = 1.44 V
DIMM = 1.87 V

This is with two different Bios (SZ10 or SZ13) for the EVGA E179 790i FTW Mobo.

Temps read (SZ10):
Core 1: 54C
Core 2: 49C
Core 3: 52C
Core 4: 52C
SPP: 64C
MCP: 75C

Temps read (SZ13):
Core 1: 59C
Core 2: 54C
Core 3: 57C
Core 4: 57C
SPP: 62C
MCP: 69C


What I'm trying to get at here is that when I OC even a little bit I really have to push the CPU Core voltage up... in the range of 1.38 to 1.4 to even get the rig into Windows. The readings in Everest correspond to 1.36ish +- 0.1.


like i told the other guy, you cant leave vtt (cpu fsb) on auto. you have to bump it up to overclock a 45nm quad. i dont consider 1.38 to 1.40 to be a high vcore. some of the other guys will. it looks like the default voltage on yours (like mine) is on the high side. do you have an e0 chip?
 
From the OCing sticky,

To start with, select a set of vcores that are kinda low and see if you can POST. How do you know where to start? Use trial and error at this point unless you know someone else’s settings to use as starting points. When in doubt, I’d recommend that you start near the bottom of the scale. Here are some rough guidelines for setting your VTT:

1.2-1.3V - for a FSB of ~400 MHz.
1.4-1.5V – for a FSB of ~420-440 MHz (exceed 1.4V at your own risk with a 45nm chip)!
1.6V – for a FSB of ~440-475 MHz - use at your own risk with a 45nm chip!

You should be aware that newer 45nm fab chips are MUCH less tolerant toward high VTT than their 65nm predecessors. Anantech published their experience frying a QX9650 with high VTT’s as an example.
 
Back
Top