• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

OCing a Q9450

N

Nicolaskl

Gawd
Joined
Oct 7, 2004
Messages
973
What are people hitting? I checked the sticky thread and only four responses in there so figured I'd ask here. I've been out of the OC game since the earlyish 939 days so I may need some advice. Currently I've got it OCed to 3.5 GHz which is the "Crazy" setting on my Rampage Formula. Rock solid so far, I started at QX9650 levels and worked up.

According to the manual it looks like those settings just adjust the FSB and the DRAM settings, no bumping the voltage anywhere? Not sure, but that's how it seems, anyone know for sure? If I'm hitting 3.5 on stock voltage it must have a decent amount left in it, right?

All I can say is, MS FSX on 9450 at 3.5 GHz and CF 4870s compared to my former rig, AMD 4000+ and 7950 GT, difference is insane. It's like a whole new game. Loving it.

What's the stock voltage run on these also, I think in my BIOS it just says "Auto" so I'm not even sure.
 
I dunno, I too am dissapointed in the lack of reports on the Q9450. I think in the face of cheap Q6600's hardly any of us enthusiast people opted for the 45nm Q9450. :(

I bought one about a month ago but can't get it past 3.0ghz. I don't think my Gigabyte P35-DS3L likes it very much. I've just bought an EVGA 750i FTW that I'm hoping will provide much better results.

Stock voltage is usually 1.25v but can vary based on the chip and it's VID.
 
Wow, 3.0 seems really low for these guys. Hopefully it's the mobo. Based off my experience so far it seems 3.7 or 3.8 on mine shouldn't be too tough, but it's also possible it's just a coincidence and 3.5-3.6 is my max. Even at 3.5 I'm pretty damn happy, I thought about just leaving well enough alone, but, you know...
 
I have my X3350 (pretty much the same chip) at 3.6 on an MSI P-35...
I have a shiny new Q9450 sitting here also for my kids machine but
I cant get the BIOS to update on his P5NE..if I ever get it to work, I can let you know how that one does...

as far as the auto voltage setting , I would be careful, you never know how much voltage its feeding the CPU....I had heard that some boards give crazy volts on auto when overclocking ..check in CPUz just to be safe...
I need 1.350 to get 3.6 on the Xeon chip.
the board has wicked droop, loaded its 1.296
 
I've had mine stable (14 hours prime) at 400x8. Had vcore at 1.17500v, in BIOS. (I was content to stop there.)

Had that OC on for about 2 months, and have turned that OC off as of a couple days ago for summer time.
 
Hmm is there any reason to get a Q9450 over a Q6600? I mean ofcourse per clock performance is much better on the Q9450 and its more power efficient and runs cooler but I mean on new egg its about $330 (though at Fry's 4th of July sale its $250) whereas the Q6600 is only $210 and under $200 through rebates. Is the extra $100 worth it or should one simply wait for Nehalem?
 
I've had a 9300, 9450, and two 6600's and my 45nm CPU's can reach higher OC's on the same board (one in my sig). Both my Q660's were SLACR G0 steppings. My 9300 performed better than my 6600, even with less cache.

If money is an issue, get the 6600. If you can jump on the Fry's deal, I'd buy the 9450.
 
Are there any differences between batch numbers and pack dates yet?

I've got an L813a719 packed on 6/5/08 and a L818a757 on 6/17/08.
I only want one, so should I use the earlier or later for better OC?
 
I'd guess that most people who planned to overclock bought Q6600's and many people who bought Q9450's didn't plan to overclock much if at all.

When I first got my Q9450 in winter I ran it at 3.6ghz for a while, but have it at stock now for the summer since it's plenty fast enough for me. My chip's VID is 1.21v, so higher than others. When overclocked, voltage was set automatically by the BIOS.
 
See sig. Most peeps with good hardware can get 3.6. I'm always unlucky with my oc's, and I hit it.

4.0 is a stretch though, even with a cherry picked chip few are getting it on air.
 
I will copy my BIOS settings later for my Q9450 later and post them here. It runs 3.6GHz at 1.19375V set in BIOS, 1.178V measured (load or no). Maybe it will help you guys out. Or maybe my chip is just ridiculous. You'll have to find out!
 
Crosshairs said:
I have my X3350 (pretty much the same chip) at 3.6 on an MSI P-35...
I have a shiny new Q9450 sitting here also for my kids machine but
I cant get the BIOS to update on his P5NE..if I ever get it to work, I can let you know how that one does...

as far as the auto voltage setting , I would be careful, you never know how much voltage its feeding the CPU....I had heard that some boards give crazy volts on auto when overclocking ..check in CPUz just to be safe...
I need 1.350 to get 3.6 on the Xeon chip.
the board has wicked droop, loaded its 1.296
Click to expand...

UPDATE:

OK, Rather than keep messing with trying to update the BIOS on the P5,I just went and swapped in an Abit IP-35 Pro .. initial results are 3.4...425X8...@ 1.200 in BIOS ..
I'm sure it has more, but I just set it there and left it for now.

The VID on this chip is really low....1.0XX something I think ..ill post it tonight when Im home again
 
all that a low VID means is that if your cpu has a vid of 1.10v, at a given vcore, say 1.30v, the cpu will run much hotter than a cpu with a vid of 1.20v. there is absolutely no data backing up wither that low vid is a better overclocker, or the other guys that say high vid is a better overclocker.
 
vanilla_guerilla said:
all that a low VID means is that if your cpu has a vid of 1.10v, at a given vcore, say 1.30v, the cpu will run much hotter than a cpu with a vid of 1.20v. there is absolutely no data backing up wither that low vid is a better overclocker, or the other guys that say high vid is a better overclocker.
Click to expand...

I'm not sure why you posted that, I didn't see anyone mention any relation to VID and overclocking.
And on my rig,with about 6 different CPU's now, I get the same temps at 1.1 stock clocks as I do at 1.6 overclocked ....high end water cooling FTW
 
All settings not listed are set to Auto. Your RAM settings may differ. My RAM is at 900MHz at these settings. Measured voltages from the BIOS are after the settings.

Multiplier: 8.0
FSB: 450
PCIE Freq: 100

CAS# Latency: 5
RAS# to CAS# Delay: 5
RAS# PRE Time: 5
RAS #ACT Time: 15

CPU Voltage: 1.19375
CPU PLL Voltage: 1.54
NB Voltage: 1.49
DRAM Voltage: 2.10
FSB Termination Voltage (VTT): 1.36
South Bridge Voltage: 1.10
Loadline Calibration: ENABLED
CPU GTL: 0.67x
NB GTL: 0.67x
DRAM Controller Voltage: DDR2_REF

CPU Spread Spectrum: DISABLED
PCIE Spread Spectrum: DISABLED

C1E Support: Disabled
Vanderpool: Disabled
Execute Disable Bit: Enabled
Max CPUID Value Limit: Disabled


MEASURED VOLTS IN BIOS:

CPU: 1.176
CPU PLL: 1.616
FSB Term (VTT): 1.296
DRAM: 2.16
NB: 1.504
SB: 1.136
SB 1.5V: 1.536

Hope this helps you guys. Some of the stuff (PLL, FSB term, GTL) I have not gone back and found my minimums, but for the CPU Voltage I did. Check my post in the C2D database thread for screenshots of that.
 
Sysjack said:
All settings not listed are set to Auto. Your RAM settings may differ. My RAM is at 900MHz at these settings. Measured voltages from the BIOS are after the settings.

Multiplier: 8.0
FSB: 450
PCIE Freq: 100

CAS# Latency: 5
RAS# to CAS# Delay: 5
RAS# PRE Time: 5
RAS #ACT Time: 15

CPU Voltage: 1.19375
CPU PLL Voltage: 1.54
NB Voltage: 1.49
DRAM Voltage: 2.10
FSB Termination Voltage (VTT): 1.36
South Bridge Voltage: 1.10
Loadline Calibration: ENABLED
CPU GTL: 0.67x
NB GTL: 0.67x
DRAM Controller Voltage: DDR2_REF

CPU Spread Spectrum: DISABLED
PCIE Spread Spectrum: DISABLED

C1E Support: Disabled
Vanderpool: Disabled
Execute Disable Bit: Enabled
Max CPUID Value Limit: Disabled


MEASURED VOLTS IN BIOS:

CPU: 1.176
CPU PLL: 1.616
FSB Term (VTT): 1.296
DRAM: 2.16
NB: 1.504
SB: 1.136
SB 1.5V: 1.536

Hope this helps you guys. Some of the stuff (PLL, FSB term, GTL) I have not gone back and found my minimums, but for the CPU Voltage I did. Check my post in the C2D database thread for screenshots of that.
Click to expand...

Thanks looking to pick up this CPU next week along with some DDR2 1000. This def helps :).
 
Crosshairs said:
I'm not sure why you posted that, I didn't see anyone mention any relation to VID and overclocking.
And on my rig,with about 6 different CPU's now, I get the same temps at 1.1 stock clocks as I do at 1.6 overclocked ....high end water cooling FTW
Click to expand...


the post before mine mentioned.

i was wondering how high end does your water cooling have to overturn the laws of physics and thermo-dynamics?
 
vanilla_guerilla said:
the post before mine mentioned.

i was wondering how high end does your water cooling have to overturn the laws of physics and thermo-dynamics?
Click to expand...

The simple answer would be you need to dissipate more heat than you generate...not really all that difficult with the right parts set up the proper way.

My loop has the capacity to handle twice the load I can generate ...Its not really rocket science ..
And the post above yours was mine,I made no mention of the correlation between VID and overclocking.....There was however a discussion going on at OCF about that exact subject....

One person had provided data that a low VID does in fact provide a better clocking chip...the discussion got quite a lot of attention, and the end result was that most people thought that he did not test enough chips to support his claim.....


I have no opinion one way or the other on that subject.
 
I've seen a few hit 3.6ghz plus. Problem is high FSB is a challenge to achieve on quad cores.
 
Sysjack, thanks. This will help me with the Q9550 I plan to get at the end of July when prices drop.
 
just just set mine at 450 fsb, 1.2v and thats it, runs great no tweaking or messing around just left it lol
 
I just started OCing my Q9450.

Currently at 8x400 with 1.25 stock voltage, RAM 5-5-5-18 @ 2.1 and +.1 into my NB I'm 100% stable.

1.2 - my last core fails immediately during Prime95
1.2125 - Prime95 crashes
1.225 - my last core fails after 15 minutes

According to Real Temp my temps @ 1.25 cap out at 48-51c and idle at 39c. At 1.225 I idle at 38c and under load it's 47-50c. So there's no real appreciable temperature difference. At that point is it even worth dropping the vcore from 1.25 to 1.225 assuming 1.225 was stable?

Now I have a few questions as I've been out of the overclocking game for awhile and things have changed a lot since I last oced my Athlon XP 2500+ to 3300+. ;)

In BIOS I choose 1.25. I am using Loadline Calibration.
PC Health status calls it 1.300 which I'm guessing is due to Loadline Calibration or is BIOS just that wrong?
Hardware Monitor calls it 1.28 when idle and 1.26 under load. I'm guessing the drop is due to vdroop?
CPU-Z calls it 1.28 when idle and calls it 1.264 which I assume is due to vdroop but it's still higher then what I set in BIOS so again I'm guessing Loadline Calibration?

Which is the correct vcore value?

I can post at 8x450 @ 1.25 vcore but I immediately BSOD when it starts to load windows. Boosting my vcore to 1.275 allows me to not BSOD but the system locks up instead. Boosting the NB more doesn't do anything. So am I safe to assume that more vcore may stabilize me at 3.6?

Thx.

-V
 
Vermillion said:
I just started OCing my Q9450.

Currently at 8x400 with 1.25 stock voltage, RAM 5-5-5-18 @ 2.1 and +.1 into my NB I'm 100% stable.

1.2 - my last core fails immediately during Prime95
1.2125 - Prime95 crashes
1.225 - my last core fails after 15 minutes

According to Real Temp my temps @ 1.25 cap out at 48-51c and idle at 39c. At 1.225 I idle at 38c and under load it's 47-50c. So there's no real appreciable temperature difference. At that point is it even worth dropping the vcore from 1.25 to 1.225 assuming 1.225 was stable?

Now I have a few questions as I've been out of the overclocking game for awhile and things have changed a lot since I last oced my Athlon XP 2500+ to 3300+. ;)

In BIOS I choose 1.25. I am using Loadline Calibration.
PC Health status calls it 1.300 which I'm guessing is due to Loadline Calibration or is BIOS just that wrong?
Hardware Monitor calls it 1.28 when idle and 1.26 under load. I'm guessing the drop is due to vdroop?
CPU-Z calls it 1.28 when idle and calls it 1.264 which I assume is due to vdroop but it's still higher then what I set in BIOS so again I'm guessing Loadline Calibration?

Which is the correct vcore value?

I can post at 8x450 @ 1.25 vcore but I immediately BSOD when it starts to load windows. Boosting my vcore to 1.275 allows me to not BSOD but the system locks up instead. Boosting the NB more doesn't do anything. So am I safe to assume that more vcore may stabilize me at 3.6?

Thx.

-V
Click to expand...


Back the multiplier off to 6x and find out your max fsb first. Did you do that?
 
Nicolaskl said:
All I can say is, MS FSX on 9450 at 3.5 GHz and CF 4870s compared to my former rig, AMD 4000+ and 7950 GT, difference is insane. It's like a whole new game. Loving it.
Click to expand...

What kind of FPS are you getting with this setup? Im in the process of building a rig for FSX.. its nerve wrecking though..

FSX basically trumps any modern hardware due to poor coding, and intense on the fly rendering..
 
bumped mine to 1.23v, 475x8 ran prime95x4 for 12 hours no issues, 51deg max

ram is @ 2.0v
 
OC my Q9450 last night from 3.1ghz to 3.6ghz with 1.3v, and it's stable in prime95. My temps around 45 idle and 55-60 @ load (depending on the game). These processors are amazing.

using the OCZ OCZTVEND2 120mm Rifle cpu HSF...
 
ive got mine running @ 3.4 at 1.136 voltage, keeping it here just for long term durability - i may try to push it higher next week. I don't know. :p

I've had a e8400, another q9450 and finally this one. Just remember guys to keep the volts under 1,4, the degrading thing does happen ive witnessed it on my e8400, so just looking out for future 45nm owners
 
silik said:
ive got mine running @ 3.4 at 1.136 voltage, keeping it here just for long term durability - i may try to push it higher next week. I don't know. :p

I've had a e8400, another q9450 and finally this one. Just remember guys to keep the volts under 1,4, the degrading thing does happen ive witnessed it on my e8400, so just looking out for future 45nm owners
Click to expand...

How high did you have the core voltage and for how long? What were your temps reading on the cores?
 
my e8400 was ran at 1.45 for a few benches, not very long, then down to 1.4 for 4.2ghz for more benches, then i had to leave it at 1.375 to get 4.0 stable, this was about for a week.

Roughly after I noticed it was no longer prime stable, had to up the vcore a bit more to 1.39.

I'm not sure if it happens to all the chips because i know people that still do 1.4+ on the penryns, but just sharing my experience, also mine was under water, the highest temp it ever reported was 48 under load @ 1.45 using real temp.
 
With temps like that, it sure seems like it was dielectric breakdown. Pretty scary because you didn't seem to really push it that hard. Thanks for the info.
 
You must log in or register to reply here.
Back
Top