Christmas Eve Overclocking Log

Pneumonic

Pneumonic

Pneumonic

Soldato
Joined
19 Oct 2007
Posts
3,246
Location
Suffolk
Ok... well I have some time on my hands this afternoon and this evening so I thought I would post a Log of my latest Prime95 8 hour overclock attempt...

So far 3.906Ghz has been the highest clock at ambient temps and a 4.023Ghz run was achieved with a little help from mother nature and an open window :) Tonight am at my parents house for christmas tomorrow and they have the great technical advantage of a bloody cold conservatory... So I thought lets see what can be achieved ;)

The aim of the session and log is a) I am bored! b) hopefully some people will find this interesting c) to get to 4.1Ghz on air! :) 4.1Ghz maybe a little ambitious but you dont get anywhere without trying!

So The Setup:

Basically no different from any of my previous runs... for those that arent aware the hardware is as follows:

* Asus P5K-E
* G0 Q6600 - Lapped
* 4*1 Gb GeIL ULL
* Thermalright Ultra 120 - Lapped
* 3 Delta Delta EFB1212SHE 120x38mm Triple Blade (3700 RPM, 142 CFM, 53Dba)

I get asked a lot about how loud the system is... at idle no louder than your average desktop.. all my fans are speedfan controlled and run around 30% which is around 1000RPM on the fans... but for those that are interested I have put together a little video of the fans stepping up through the RPM :) low quality version can be found here and "higher" quality here.. The camera is around 30cm from the case

Temperatures:
The biggest factor in the clock is going to be the temperatures that the cores will reach, so hopefully tonight will be cold enough :)

Temps.jpg


Here are the current temperatures at idle, I am guessing the ambient temperature at the moment is around 8c my system temperature usually is around +4/5c over ambient and my little widget thing is telling me that its currently 8c locally

The Clock:
To achieve 4.1GHz I will run a *9 Multiplier on the CPU and a FSB of 456mhz which will give a clock speed of 4.104Ghz

More to follow shortly am now off to have a play around and see what voltages will be requried
 
Last edited:
Update:

Well I have spent the last hour playing around...

First attempt @ 4.1Ghz I started by using settings that I know are stable for 32m SuperPi just to give myself a bench mark of how much more a prime95 test was going to take

BIOS as Below:
Vcore BIOS: 1.575v
Vcore Windows (Idle): 1.552v
Vcore Windows (Load): Machine Crashed Before I could look at CPUz after I hit the prime go button :)

I upped the vCore volts all the way to 1.625v and everything had the same effect after a play around with some other BIOS settings... PLL etc.. still no luck
Problem here I think is that between 450mhz and 467mhz I know that the motherboard has some serious stability issues so could be the main problem :(

Best I got @ 4.1Ghz was 25 mins of prime :) temps loading to around 62/62/56/56 @ Vcore 1.625 (BIOS) 1.600 (Idle) 1.584 (Load) shame because temps were looking nice... damm and blast my motherboard


So overclock for tonight looks like it may be limited to 449*9 which comes out at 4041 still an improvement but not as much as was first hoped for

Culprit Identified??

Well after having reboots at both 4041 and 4032 I decided to load up my BIOS settings for the stable 4023 run and just run a prime on that (got to love saved BIOS profiles) just in case I had missed a setting... Well 5 mins in I get another reboot all rather strange :(

So after a few choice swear words only thing I could think that was different was the fact that I had my WiFi connection enabled... so off that went... back into windows and a quick prime on 4023 settings and everything seemed stable... past the 5 mins where it was normally failing.... so back to the BIOS and FSB up one notch... system is now 30 mins in to a prime @ 4032... will give it a couple of hours and then notch it up once more for a full run @ 4041

Update 2:

Well have run Prime95 for just over an hour @ 4032 so time to up the clock... screeny below:

4032.jpg


Settings as Follows For that Run:

Vcore: 1.585v (BIOS), 1.560/8v (Idle), 1.544v (Load)
NB: 1.55v
SB: Auto
PLL: 1.6v
RAM: 896mhz @ 5-5-5-23 very lose to avoid a prime fail due to RAM problems
FSB: 448mhz

Update 3:

Well given that I identified that the WiFi was making the board was unstable have now gone through the BIOS and disabled everything that I can think of! Onboard, LAN, Sound, WiFi and Serial stuff... seems to have made a difference now have what looks like a stable Prime @ 4086... looks like 4100 could be back on :)

Update 4:

Well was back on for all of 12mins until the machine rebooted hehe!! Might go nick the brothers Asus MF board :) anyways back to being sensible.. up to 4041mhz and leave it running for a bit
 
Last edited:
Update 5:

Well disabling everything has sorted stability issues at up to 449FSB but unfortunatley the motherboard still gives out at anything over 450FSB which is rather a shame because I am certain that the chip can go faster :) But 4041 I am still very happy with leaving it to Prime now will post some temperatures and settings once I have it going for an hour +

Update 6:

Well rather annoyingly it seems that 449FSB is unstable as well... so back down to 448FSB, got a good hour on prime earlier so *fingers crossed* this will hold out for 8 hours... still an improvement on 4023 and I know for certain now I am at the limit of the motherboard... really looking forward to the new yorkfields and the next gen of motherboards now :) see how far Air can go

As promised earlier.. some temps for you :) 57/57/51/51 @ 30 mins into the 4032 run... same BIOS settings as the hour run earlier

Update 7:

Still going strong temps are holding as above and is nearly 1/2 through the first cycle of 8 - 64K another couple of hours till it goes through 64K and back to the start if it holds that long should be good for the whole 8 hours

Update 8:

Well 3 1/2 hours in an machine reboots after some discussion with W3bbo (thank you very much!!) I have a better idea of how to stabalize the FSB.. but that will have to wait for another day as I have to be up early to stuff a bird... unfortunatley one of the feathered variety...

So for tonight I will leave you with some Load temps @ stock :)

Stock.jpg
 
Last edited:
Good luck with it. Not worth using a budget low profile graphics card if you have one? So it's not heating everything up.

Also, :eek: at the loudness of the machine in that HQ video at around 30 seconds left, practically hoover levels. Are you aiming for Prime95 stable or SuperPI stable?
 
flibby said:
Good luck with it. Not worth using a budget low profile graphics card if you have one? So it's not heating everything up.

Also, :eek: at the loudness of the machine in that HQ video at around 30 seconds left, practically hoover levels. Are you aiming for Prime95 stable or SuperPI stable?
Click to expand...

Prime95... am SuperPI stable @ 4.1 :)
 
Definitely a very good clocking chip! I can get mine to boot at 4ghz but requires 1.65v BIOS and it certainly isn't stable even under water! Might attempt to sell this on an auction site and get another and try that. VID on mine is 1.265v, just doesnt like high clocks!
 
I have just got the P5K-E

What PLL voltage
NB
SB

also what bios?

I am about for a hour so ill see if i can help with the settings if you think board may be the concern ;)
 
Pneumonic said:
would leave them all at auto unless you are going over 3.6 Ghz clock you shouldnt need to touch them.. BIOS i am on the beta 0701 I think
Click to expand...

I was going to see what you had see if i can help you :p Nice clocking tho and we all love threads like this. hehe

May i ask whats the Max FSB you achieve with your board/chip. Multi at 6 highest FSB. My board and CHIP maxes 495FSB.

Does 701 bios have the voltage damper option as i know they took it off 901 bios?
 
4Qman said:
I was going to see what you had see if i can help you :p Nice clocking tho and we all love threads like this. hehe

May i ask whats the Max FSB you achieve with your board/chip. Multi at 6 highest FSB. My board and CHIP maxes 495FSB.

Does 701 bios have the voltage damper option as i know they took it off 901 bios?
Click to expand...

yes has the voltage dampner :) my best friend hehe otherwise I get huge vdrop/vdroop

I can get 497FSB at a *6 muli stable then up to around 503FSB its touch and go :) and as I said earlier 450-467 is also touch and go
 
Not bad mate, ill give 701 a try. Mine is stable 495 with PLL etc maxed on 601 bios. Add cooling to mosfits near CPU if you havent already as they got hot. Ill Vmoding when i get time 2mra so see how it goes
with some more juice.

good luck keep updated man ;)
 
4Qman said:
Not bad mate, ill give 701 a try. Mine is stable 495 with PLL etc maxed on 601 bios. Add cooling to mosfits near CPU if you havent already as they got hot. Ill Vmoding when i get time 2mra so see how it goes
with some more juice.

good luck keep updated man ;)
Click to expand...

I have thought about the volt mod... need to go and read over at extreme about getting a higher stable FSB as I believe there is a mod for that... vcore is fine atm.. is the darn FSB that isnt :(
 
Looks like I can get to 4.005ghz, takes 1.6v BIOS, 1.56v Windows and it goes to 1.52v under load.
Its not prime stable however its be fine for general use and 3DMark06 tonight, will have to play around with PLL if its available on P5K Deluxe!

What does the PLL voltage refer to then and what does it do exactly? :)

(Stolen from someone who was running 7x500)

CPU Voltage Reference: x0.63
CPU PLL Voltage: 1.50
DRAM Voltage: 2.45 (added 0.05v just for precautions when setfsb is limited by mem clocks)
FSB Termination Voltage: 1.40
North Bridge Voltage: 1.70
North Bridge Voltage Reference: AUTO
South Bridge Voltage: 1.20

All of those are available on the P5k, not changed any of those above though, will have to have a read around about them.
 
Last edited:
PLL is a large factor with high FSB increments however having a CPU with the balls to give high FSB is the main factor.

As w3bbo said extreme high PLL will get those last FSB increments. I read Intel say 1.6v PLL is 24/7 limit for safe running and long life. Not that it bothers me.
 
Just watched your video with the delta givin it sum .. LOL :p BTW, have you heard 6 of these givin it sum on a 120.3 radiator? My wife has and she told me to ditch em ... fast!!! :D

Santa has just deployed his sack full of goodies so tis time for a good couple hours sleep methinks ......
 
4Qman said:
I read Intel say 1.6v PLL is 24/7 limit for safe running and long life. Not that it bothers me.
Click to expand...

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=3156&p=5

CPU PLL Voltage: Maximum of 2.78V. Out of all the voltages the user can manipulate this one is by far the most dangerous. Maximum vCPUpll, as established by Intel, is 1.60V (default for this board) making 2.78V a whopping ~75% over specification! (As an aside, this would be the equivalent of subjecting your 65nm CPU to a core voltage of over 2.5V). Exercise extreme caution when utilizing higher values as setting this value too high can result in the CPUs "losing cores" after being subjected to voltage in excess of ~2.0V. The good news is that we did not see an increase in overclocking potential with voltages above 1.68V.
Click to expand...
 
You must log in or register to reply here.
Back
Top Bottom