Q9450 OC / Temps / Settings

[truehighroller]

Mine died after running it at 3.8Ghz for about 2 months so I think they suck now. Bad thing is I lapped mine and now can't do anything with it but use it for a paper weight. Worst $300 I have ever spent..

[RunawayPrisoner]

Yeah... I'd think you should not run it at extreme voltages. Just... be moderate. This chip is taking in a lot more voltages than its stock at higher clock frequencies.

And anyway, 9880 3DMark Vantage.

[crazyea]

I did 4 GHz on the new Q , Cant do that on Q66

Hmm, was the opposite for me. I still need more time to figure this q9450 chip out.



Runaway, What kind of settings, and is it prime stable?

[RunawayPrisoner]

Prime for a few hours... temps are very crazy (average 75's to 80's) so I didn't want to keep it up. It's stable enough to play games and encode movies, though, so I guess it should be fairly stable, or barely stable.

[003]

Yeah... I'd think you should not run it at extreme voltages. Just... be moderate. This chip is taking in a lot more voltages than its stock at higher clock frequencies.

You know what? Screw you and your ridiculously high overclock with an obnoxiously low voltage.

Just you wait until I get a screenshot of mine doing Prime95 at 3.9GHz with 1.525v and CPU-z open!!

[truehighroller]

You know what? Screw you and your ridiculously high overclock with an obnoxiously low voltage.

Just you wait until I get a screenshot of mine doing Prime95 at 3.9GHz with 1.525v and CPU-z open!!

Your gonna fry your chip within a month like that. Mine fried with just 1.26v~ @ 3.8Ghz and 62 Celcius at full load, for hours of prime that is..

[003]

Then you fried it with something other than vcore, or it was faulty from the get go.

[Zigosity]

Then you fried it with something other than vcore, or it was faulty from the get go.

Agreed.

I'd be interested to know what your VTT\PLL was set at...

-Zigosity

[003]

What is the maximum safe VTT and PLL for 45nm chips btw? I use 1.58 PLL and 1.38 VTT for 3.9GHz. Is that all right?

[RunawayPrisoner]

PLL should not exceed 1.70v, and under no circumstances should you use more than 1.4v of VTT. I really think we have to stress that Anandtech killed their test QX9650 with 1.45v through VTT. And...



Well... not like I didn't do it. That chip was only under 1.38v of VTT, though.

[003]

Ok...






PS: I know my vantage score sucks, but just concentrate on the CPU score. This will change when I get my dual 4870X2s

[Zigosity]

I got a higher score at 3.6 o_O. It's a couple pages back, 5696.

Vista vs. XP, maybe? Heh.

And PLL shouldn't go above 1.6 for 45nm, as far as I've heard, 1.4 for VTT is correct though.

-Zigosity

[003]

I got a higher score at 3.6 o_O. It's a couple pages back, 5696.

Vista vs. XP, maybe? Heh.

And PLL shouldn't go above 1.6 for 45nm, as far as I've heard, 1.4 for VTT is correct though.

-Zigosity

err... Vantage does not run in XP

So I'm safe with my volts then.

But anyway... I have no idea why you are scoring higher. Err.. obviously you have a better GPU setup than I currently have, but... ??

EDIT:
Doh! You were comparing scores from 3DMark06 and Vantage, you can't do that

[003]

Runaway, what FSB strap to NB are you using?

Anybody know if it is possible to flash Rampage Formula to P5Q?

[RunawayPrisoner]

I set it at Auto? But I think it is using the 333MHz strap.

By the way, the 4GHz screenshot was from way back there, when my second Q9450 (another golden) hit 4GHz with a P5E-VM HDMI. So far... I have had two chips at 4GHz with almost 1.5v, and one chip... unknown capacity, but I don't expect much more, since 3.6GHz is under 1.2v, and I'm kinda happy.

...I think I'll send the P5Q back for a third P5E-VM HDMI...

[Zigosity]

err... Vantage does not run in XP

So I'm safe with my volts then.

But anyway... I have no idea why you are scoring higher. Err.. obviously you have a better GPU setup than I currently have, but... ??

EDIT:
Doh! You were comparing scores from 3DMark06 and Vantage, you can't do that

No I wasn't ;_;. I was talking to Runaway! I shoulda quoted his post, my bad. I don't use Vista\Vantage at all.

-Zigosity

EDIT: Oh, and as for safe with those voltages, I wouldn't exactly call 1.5+ vcore safe... Apparently max safe for that is 1.37-1.4.

[RunawayPrisoner]

Yeah, it was a little bit slow. Probably something else killed something... but it was indeed at 4GHz. Or... that was a problem with the board itself.

[mach82]

I really think we have to stress that Anandtech killed their test QX9650 with 1.45v through VTT.

nope, they assume VTT was the culprit. Yet there's no proof for that. There are a number of possibilities in this case, including a really bad CPU voltage (vcore) spike, which might have caused the death of the chip. So the whole consensus of max VTT is based on the death of one chip which was seemingly caused by excessive VTT. Sweet ...

[truehighroller]

Agreed.

I'd be interested to know what your VTT\PLL was set at...

-Zigosity

gTL references = .63 and .63 then my vPLL = 1.56 , CPU voltage was 1.26 bios actual 1.256~ No droop in other words it stayed around there all the time... I don't know what the vtt is even. I don't think I had that option to change.

[Slovnaft]

holycraponyourmomsshoes runaway, i'm pretty sure i have the same batch.
just got it in from newegg RMA blesstheirsouls and seated it in my POS giga mobo last night and got 3.7 at 1.23v (my board has some killer vdroop)
i'm not gonna try and push it any further until I get it on BlackOps which is coming, again, courtesy of godblessemneweggRMA, for my mistake of a mobo.

[truehighroller]

gTL references = .63 and .63 then my vPLL = 1.56 , CPU voltage was 1.26 bios actual 1.256~ No droop in other words it stayed around there all the time... I don't know what the vtt is even. I don't think I had that option to change.

No one, are these not safe?

[RunawayPrisoner]

nope, they assume VTT was the culprit. Yet there's no proof for that. There are a number of possibilities in this case, including a really bad CPU voltage (vcore) spike, which might have caused the death of the chip. So the whole consensus of max VTT is based on the death of one chip which was seemingly caused by excessive VTT. Sweet ...

Intel themselves said so to Anandtech... And there are indications all over the place. If vCore doesn't do it, then VTT has gotta be it. Remember, default VTT on these chips is very low. 1.10v (apply to all chips). You pump in 1.3v VTT, and it's already as good as running the chip at 1.4-1.5v of vCore. Hence why up to 1.5v of vCore, it's fine, but excessive VTT might not be fine. Anandtech was running at 1.45v, a whooping .35v increase from 1.1v. Assuming the chip's default VID to be at 1.2v (you know this is not true, it's usually 1.25v - 1.30v), then it's as good as running at 1.55v vCore (assuming they have a good chip). Now tell me how that is not lethal.

gTL references = .63 and .63 then my vPLL = 1.56 , CPU voltage was 1.26 bios actual 1.256~ No droop in other words it stayed around there all the time... I don't know what the vtt is even. I don't think I had that option to change.

VTT is FSB voltage. Its lowest value is 1.20v on most boards. When a 45nm processor is detected, decrease that by .10v. But... just to be on the safe side, never go beyond 1.40v with that voltage even in the BIOS.

holycraponyourmomsshoes runaway, i'm pretty sure i have the same batch.
just got it in from newegg RMA blesstheirsouls and seated it in my POS giga mobo last night and got 3.7 at 1.23v (my board has some killer vdroop)
i'm not gonna try and push it any further until I get it on BlackOps which is coming, again, courtesy of godblessemneweggRMA, for my mistake of a mobo.

Yeap, this has been a good batch so far. And who was it that asked for a 24-hour small FFT prime screenshot?

[truehighroller]

Intel themselves said so to Anandtech... And there are indications all over the place. If vCore doesn't do it, then VTT has gotta be it. Remember, default VTT on these chips is very low. 1.10v (apply to all chips). You pump in 1.3v VTT, and it's already as good as running the chip at 1.4-1.5v of vCore. Hence why up to 1.5v of vCore, it's fine, but excessive VTT might not be fine. Anandtech was running at 1.45v, a whooping .35v increase from 1.1v. Assuming the chip's default VID to be at 1.2v (you know this is not true, it's usually 1.25v - 1.30v), then it's as good as running at 1.55v vCore (assuming they have a good chip). Now tell me how that is not lethal.

I think I had



VTT is FSB voltage. Its lowest value is 1.20v on most boards. When a 45nm processor is detected, decrease that by .10v. But... just to be on the safe side, never go beyond 1.40v with that voltage even in the BIOS.



Yeap, this has been a good batch so far. And who was it that asked for a 24-hour small FFT prime screenshot?

AI Overclock tuner: manual
CPU Ratio Setting: 8
FSB Frequency: 475
FSB Strap to North Bridge: 333
PCI-E Frequency: 101
DRAM Frequency: DDR2-1141MHz
DRAM CLK Skew on Channel A1: auto
DRAM CLK Skew on Channel A2: auto
DRAM CLK Skew on Channel B1: auto
DRAM CLK Skew on Channel B2: auto
DRAM Timing Control: manual

1st Information :
CAS# Latency: 5
DRAM RAS# to CAS# Delay: 5
DRAM RAS# Precharge: 5
DRAM RAS# Activate to Precharge: 15
RAS# to RAS# Delay : 3
Row Refresh Cycle Time: 55
Write Recovery Time: 5
Read to Precharge Time: 3

2nd Information :
READ to WRITE Delay (S/D): 7
Write to Read Delay (S): 3
WRITE to READ Delay (D): 5
READ to READ Delay (S): 4
READ to READ Delay (D): 6
WRITE to WRITE Delay (S): 4
WRITE to WRITE Delay (D): 6

3rd Information :
WRITE to PRE Delay: 14
READ to PRE Delay: 5
PRE to PRE Delay: 1
ALL PRE to ACT Delay: 6
ALL PRE to REF Delay: 6
DRAM Static Read Control: Disabled
DRAM Read Training: auto
MEM. OC Charger: Enabled
AI Clock Twister: Stronger
AI Transaction Booster: Manual
Common Performance Level 8
Pull-In of CHA PH1: Enabled
Pull-In of CHA PH2: Enabled
Pull-In of CHA PH3: Enabled
Pull-In of CHA PH4: Enabled
Pull-In of CHB PH1: Enabled
Pull-In of CHB PH2: Enabled
Pull-In of CHB PH3: Enabled
Pull-In of CHB PH4: Enabled

CPU Voltage: 1.25xx
CPU GTL Voltage Reference (0/2): 0.630
CPU GTL Voltage Reference (1/3): 0.640
CPU PLL Voltage: 1.56
FSB Termination Voltage: 1.28
DRAM Voltage: 2.02 actual is 2.1
NB Voltage: 1.28
NB GTL Reference: 0.630
SBridge Voltage: 1.20
PCIE SATA Voltage: 1.60

Load Line Calibration: enabled
CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled
CPU Clock Skew : normal
NB Clock Skew : normal

Advance CPU Settings
CPU Ratio Setting: 8
C1E Suppport: Disabled
Max CPUID Value Limit: Disabled
Intel® Virtualization Tech: Disabled
Vanderpool Technology: Disabled
CPU TM Function: disabled
Execute Disable Bit: disabled

That is what I had her set to 24/7 temps at 62 Celcius max at full load. I think it was like 35 Celcius max at idle.

[SuporterPoli]

@RunawayPrisoner: you do know that both vcore and vtt(vfsb) have the absolute maximum value at 1.45?
1.3v vtt is not dangerous.Over 1.4 and you are in cpu killing territory. Also if your board overvolts and you put in bios 1.4 or less but the real voltage is much higher, then you can assume who killed your cpu.

[mach82]

Intel themselves said so to Anandtech... And there are indications all over the place. If vCore doesn't do it, then VTT has gotta be it. Remember, default VTT on these chips is very low. 1.10v (apply to all chips). You pump in 1.3v VTT, and it's already as good as running the chip at 1.4-1.5v of vCore. Hence why up to 1.5v of vCore, it's fine, but excessive VTT might not be fine. Anandtech was running at 1.45v, a whooping .35v increase from 1.1v. Assuming the chip's default VID to be at 1.2v (you know this is not true, it's usually 1.25v - 1.30v), then it's as good as running at 1.55v vCore (assuming they have a good chip). Now tell me how that is not lethal.

Did you expect Intel say to anandtech, that exceeding nominal values (OC) is safe? That would simply encourage more people to OC and hence would also mean less money for Intel as people would OC instead of buying new CPUs. Intel and anandtech can just assume what was the cause of death of the CPU and Intel took their chance to clearly state that OC is bad and if you do it you may just kill your CPU. Truehighroller had his chip fry with a very low vcore of 1.26V. Does it meant that 1.26V can kill a 45nm chip? Or perhaps voltage regulators just can't work correctly when CPUs are heavily OCed. Some mobos are just not ready to work with quads at 4 GHz, which leads to mobos' failures and consequently dead chips.

You can also compare this situation to RAM makers. The standard voltage for DDR2 800 is 1.8V, yet some manufacturers would rate their sticks for max 2.4V, which is 33% higher than the standard value. Some of the sticks have been running such for 2 years and are still fine. On the other hand some sticks would die with 2V. Does it mean that 2V can kill DDR2? There's much more behind the phenomenon than just stating that such voltage is safe and such is not. If a given mobo has poor voltage regulators, then even safe voltage may cause a chip/stick to die. I've had my 45nm quad run a few times with VTT of 1.5V (overall about 30h) and it's still fine. I've also had my D9GMH stick killed while running with an EVGA 680 @2V within 24 hours, though it had run fine with my Asus P5Be @ 2.2V for a year. One incident can't be an oracle for all cases. I bet there are a number of people whose 45nm chips have been working with VTT>=1.45 and the CPUs are still fine.

Now look at the above settings of truehighroller - can you see any voltage that could lead to his chip being killed? I see none. If we were working for anandtech, we could state, that 1.26V can kill a 45nm chip and everyone would be afraid of setting more than 1.25V