Q9450 OC / Temps / Settings

[kiikkuja]

vtt is fsb term. right?

I believe so.

[Justintoxicated]

justintoxicated, i didnt know you weren't raising your VTT volts, head up to 1.35 in bios (1.34v actual) and your proc should easily get to 3.6 seriously the vtt and 1.6v on the NB will give you that ability to do at least 460 fsb with this board.

i posted a gtl chart over in the x48 thread for you a lil while back, that should help with changing the VTT up

Unfortunately that is not the case as I already tried these settings even played with some different GTLs settings acording to other charts my probing the MB and getting the same percentages...

It seems to me I can get away with having my GTL's set incorrectly if I boost Vtt voltages, the more Vtt the more headroom I have for not having the correct GTL. I WAS increasing NB CPU and Vtt but have been working on lowering them for the past week. Unfortunately GTL charts are only somewhat helpful every proc mb may be different. GTL charts should allow you to make a good hypothesis abotu your own setup though! My chip does not match other GTL charts at all.

But hey check it out now! I'm 3200 mhz stable without adding voltage to anything else! Vtt is now at 1.100 with some nice GTL adjustments! leaving them at default levels I have to add a bunch of CPU and Vtt and maybe (dont remember exactly) NB voltages.

Now the next step is to lower GTL's until I'm not stable once I get these as low as possible I will calculate percentages and try for higher clocks keeping them set at this ratio.

Its my new hypothesis that when people switch boards and achieve much higher overclocks (under 500 FSB wise at least) that the MB is simply adjusting their GTL's closer to the CPU's needs. At least to some extent. I do think the DFI board is very capable esp under water, but the Auto GTL settings are no where close to what my CPU needs to achieve higher FSB.

GTL's are KEY!

I have to get to work but I will explain my GTL hypothesis when I get to work.

[Slovnaft]

have you tried setting everything to auto?
Initially I found the same thing with my board. I set all the voltages myself and tried ocing, but it would freeze in windows boot at 410. So I decided to work backwards, leaving all my voltages on auto, seeing how high i could get with that, then manually modding key voltages to get higher. That got me 480 fsb on auto and 490-95 with some NB and cpu tweaks.

[Justintoxicated]

have you tried setting everything to auto?
Initially I found the same thing with my board. I set all the voltages myself and tried ocing, but it would freeze in windows boot at 410. So I decided to work backwards, leaving all my voltages on auto, seeing how high i could get with that, then manually modding key voltages to get higher. That got me 480 fsb on auto and 490-95 with some NB and cpu tweaks.

Yes it is completely unstable on auto. It would not even boot over 3200 mhz even with only GTL's on auto and all other voltages pumped at several different levels. I think whatever the bio is doing on the DFI boards are justnot optimal settings for my CPU. Could be why some people change boards and get more OC and others change boards and it gets worse. I'm going to look for a GTL range that works with stock voltages then set GTL's accordingly as Vtt increase is needed requires me to manualy probe the MB with a fluke. It's a huge PITA. Actualy it would be really helpful if smeone else with a C1 stepping and a decent overclock CPU could probe their MB's GTL's and Vtt and calculate some percentages. Then again they may very well vary from board to board ...

[Earzz]

vtt is fsb term. right?

I believe so.

Yes 100% sure.

[diablos]

Hi mate,
Do you know this batch?

L823B126

thanks

[freyesm]

Guys, I'm trying 450x8, but no luck, the max I managed was 435x8 (24/7 on air). Got 440x8, but isn't 100% stable.

Using P5E with P5E Deluxe 0205 BIOS.

Anyone got better results on P5E ?

Fernando

[skinnee]

Here is where I'm at thus far...

http://valid.x86-secret.com/show_oc.php?id=412137

Running MemTest on 1840 8-8-8-18 2T presently, we'll see how far CAS8 will take me. I should have quite the write-up on the G.Skill DDR3-1600 CL7-7-7-18 (F3-12800CL7D-4GBHZ) kit when all is said and done. CAS7 bottomed out at 1780.

[kiikkuja]

Here is where I'm at thus far...

http://valid.x86-secret.com/show_oc.php?id=412137

Running MemTest on 1840 8-8-8-18 2T presently, we'll see how far CAS8 will take me. I should have quite the write-up on the G.Skill DDR3-1600 CL7-7-7-18 (F3-12800CL7D-4GBHZ) kit when all is said and done. CAS7 bottomed out at 1780.

How much vcore did that require? It seems my q9450 does like vcore too much I have to pump over 1.4V to get near 4Ghz. 3.6 or under for me it seems as I don't want to run anymore than 1.375Vcore through my chip.

[Slovnaft]

I don't know much about later batches, but I know L815 and L816 both like lololow voltages. For instance, I'm stable right now at 3.9 with 1.28 vcore. But interestingly enough in order to boot into 4ghz i need around 1.5v even 1.55v, and even then it's not guaranteed. I think it's because I haven't tweaked many other voltages, as my board responds poorly to custom volts. But I would imagine that later batches may lend themselves equally to lower volts, possibly even to the point that they may not clock as well.
and @freyesm, I'm on P5Q3 if that makes any difference. I'm quite happy with my Asus P45--Q9450 combo.

[InfinityNX]

Hi guys, I'm having a problem with my Q9450.

I have it on water, with a GTX II 360 rad, dtek fuzion v1, distilled water + nuke

Using a Maximus Formula

My core temps, all on stock, on idle, runs at a whopping 40C, which is pretty high. (Using Real Temp)

The weird thing is that my CPU temp, from bios, never goes above 30c

Anything wrong with this?

[kiikkuja]

Hi guys, I'm having a problem with my Q9450.

I have it on water, with a GTX II 360 rad, dtek fuzion v1, distilled water + nuke

Using a Maximus Formula

My core temps, all on stock, on idle, runs at a whopping 40C, which is pretty high. (Using Real Temp)

The weird thing is that my CPU temp, from bios, never goes above 30c

Anything wrong with this?

It doesn't sound too bad. My proc runs around 36C idle. What's your ambient?

[Slovnaft]

What voltage does cpuz read?
vcore to temp is a quadratic relationship, so if your board is setting it too high you'll get unreasonable temps.
40C isn't horrible for idle, but mine are in low 30s on air.
Also keep in mind the thermal diodes weren't made to give entirely accurate idle temps but to provide for thermal throttling at load.
what are your load temps?

[InfinityNX]

@kiikuja

I'm not sure about my ambient temps, but I'm in an air conditioned room.

@slovnaft

cpuz reads 1.136V on core voltage.

[Slovnaft]

my that's unusual.
that's a pretty low vcore. I'm on 1.28 (which is still low) and my idle is low 30s.
again, what's your load temp? get a burn test load temp and a prime load temp.

you could try re-seating your water block with less TIM.

[JohnnyV]

Perhaps your TJ max is set at 105c? You can check that under settings in Realtemp. One can easily "cheat" when showing off temps just by changing this.

[SNiiPE_DoGG]

Unfortunately that is not the case as I already tried these settings even played with some different GTLs settings acording to other charts my probing the MB and getting the same percentages...

It seems to me I can get away with having my GTL's set incorrectly if I boost Vtt voltages, the more Vtt the more headroom I have for not having the correct GTL. I WAS increasing NB CPU and Vtt but have been working on lowering them for the past week. Unfortunately GTL charts are only somewhat helpful every proc mb may be different. GTL charts should allow you to make a good hypothesis abotu your own setup though! My chip does not match other GTL charts at all.

But hey check it out now! I'm 3200 mhz stable without adding voltage to anything else! Vtt is now at 1.100 with some nice GTL adjustments! leaving them at default levels I have to add a bunch of CPU and Vtt and maybe (dont remember exactly) NB voltages.

Now the next step is to lower GTL's until I'm not stable once I get these as low as possible I will calculate percentages and try for higher clocks keeping them set at this ratio.

Its my new hypothesis that when people switch boards and achieve much higher overclocks (under 500 FSB wise at least) that the MB is simply adjusting their GTL's closer to the CPU's needs. At least to some extent. I do think the DFI board is very capable esp under water, but the Auto GTL settings are no where close to what my CPU needs to achieve higher FSB.

GTL's are KEY!

I have to get to work but I will explain my GTL hypothesis when I get to work.

you expecting magic when there is very little there.

I can do 400x9 with 1.1v vtt all day long, but if you EVER want to be stable at 440 plus your going to need at least 1.3v if not higher to achieve this.

I have been working with this board, overclocking quads, since its release (almost a year now), I know exactly how it behaves at different FSB levels and what the required settings are. Basically what I'm saying is that you are wasting your time trying to reach mid 400's with such a low vtt, and you need to tweak gtls when you have that higher voltage, otherwise none of the gtl tweaking you did carries over...

[Slovnaft]

@infinityNX

just occurred to me; what batch is your chip?
so much of the voltage-temp-stability dynamic depends on the characteristics of the proc batch. For instance my L815 can clock at lower voltage than earlier batches and maintains reasonable temps, whereas L816 can handle even lower volts, but yields higher temps.
Anyhow, drop voltage like its hot. because it is.

[InfinityNX]

@JohnnyV

TJMax is set at 95C. Is that the right setting?

@slovnaft

Load temps (loaded using OCCT) are around 50's.

Is there a way to know what batch of proc you have without the box/looking at the processor itself? Like a software-based tool (i.e. CPU-Z)

[Slovnaft]

yes 95 is right for TJMAX.


no. It's engraved on the IHS of the chip and on the factory seal of the box.
those are great load temps. Now try Burn Test. Also do the RealTemp sensor test, I think you'll find they're faulty, probably loose, and hence record high idle temps.

[skinnee]

How much vcore did that require? It seems my q9450 does like vcore too much I have to pump over 1.4V to get near 4Ghz. 3.6 or under for me it seems as I don't want to run anymore than 1.375Vcore through my chip.

I'm set at 1.3125 in BIOS, BIOS reports 1.35 and 1.296 CPU-Z reported. I'll get my batch number when I get home.

[skinnee]

yes 95 is right for TJMAX.

Not according to Intel as of IDF, 100ºC is Tjmax for Q9450.


From the DTS Sensor slidedeck presented at IDF:
45nm Desktop Dual-Core Processors
Intel Core 2 Duo processor E8000 and E7000 series - 100°C

45 nm Desktop Quad-Core Processors
Intel Core 2 Quad processor Q9000 and Q8000 series - 100°C
Intel Core 2 Extreme processor QX9650 - 95°C
Intel Core 2 Extreme processor QX9770 - 85°C

[Slovnaft]

Not according to Intel as of IDF, 100ºC is Tjmax for Q9450.


From the DTS Sensor slidedeck presented at IDF:
45nm Desktop Dual-Core Processors
Intel Core 2 Duo processor E8000 and E7000 series - 100°C

45 nm Desktop Quad-Core Processors
Intel Core 2 Quad processor Q9000 and Q8000 series - 100°C
Intel Core 2 Extreme processor QX9650 - 95°C
Intel Core 2 Extreme processor QX9770 - 85°C

really?! wow thanks, i had no idea.

[ante_ante]

I bought a Q9450 yesterday. Batch L819A770 and right now i´m at 400x8 with 1.18v in bios. Going to try more later when i get home.

[JohnnyV]

That looks promising Turbo. My chip seems to like volts like a baby likes his bottle of milk :P

Here's the current score: