When OC'ing my P4C I hit a wall at 240 fsb...why?

[Duvie]

NO PPL!!! I am talking about heat of the NorthBridge chipset...What style of heatsink is on that chip??? The higher the fsb the more heat the chip will produce...On the IC7 mobo and I repeat there is no temp monitoring point so no one truly knows the temp short of placing a thermometer on it.....

I suspect my heatsink is not cutting it....

I don't think the cpu is robably the cause but short of major driver issues with your vid card freezes are usually related to power or heat.....


Make sure that bios is at least version 17. I could not get a prime95 stable at 280fsb until I swicthed to version 17 then it was not only fine but at a lower vcore....

 

[superHARD]

I am happy now but I would ike to hit my goal fsb.

And it is 250. But as stated above logic tells you that upping the vcore should increase your OC'abality.

 

[superHARD]

OT...kinda.

Why does upping the vcore to 1.7 with this new bios lock up the system?

Last bios it did work fine. I am confused again!

 

[Duvie]

I don't know on that one!!! I will have to try it myself....

The question may be what is the amps on the 12v line on that power supply??? also where are those voltage numbers I wanted??? I know i guy whos antec 430 was dropping to mid 11's on the 12v and it was effecting his oc and stability in the 3.5-3.6ghz range...Maybe the board has problems regulating power ot the power supply can't deliver the amps on the rail needed for these extreme overclocks....

Now my 430watt Enermax is holing fine at 11.8 under load but I know from research I had a much higher amps on the 12v rail versus the 430 watt antec..


Do me a favor disconnect all not needed fans and drives and what not and see if you have any issues...Try running one stick of ram as well and make sure the vagp is 1.6 or 1.65v set...

 

[lxie123]

just an update.. 3dmark 2003/2000 runs for hours fine when i'm oc to 3.6 @ 1.675 volts.

But prime will fail within 30 seconds.

So should I lower it back to 3.45 (@ 1.6 volts)where 3dmark AND prime95 are stable? or leave it @ 3.6

@ 3.6 all my games/ gordian knot are running fine too, only thing that won't pass is prime95

My memory divider is set to 3:2 so i'm sure it's not the memory causing me to fail prime 3.6 and pass @ 3.45

 

[Duvie]

Originally posted by: lxie123
just an update.. 3dmark 2003/2000 runs for hours fine when i'm oc to 3.6 @ 1.675 volts.

But prime will fail within 30 seconds.

So should I lower it back to 3.45 (@ 1.6 volts)where 3dmark AND prime95 are stable? or leave it @ 3.6

@ 3.6 all my games/ gordian knot are running fine too, only thing that won't pass is prime95

My memory divider is set to 3:2 so i'm sure it's not the memory causing me to fail prime 3.6 and pass @ 3.45

That is your own call, but I can tell you many well known ppl here have lost confidence in the prime95 program has being a stable in the world of determining stability...the program version 23 has gotten extremely buggy and more complexed.. I stick with version 22 and run 2 instances of that... I save my memory stressing testing to another old stand-bye the memtest....

I still use and respect what prime tells me but I do not trust it solely....My p4b 2.4@3.24ghz showed me otherwise when prime would last 12 hours and fail in many rendering and encoding test..i was very disappointed in it. However if we can trust it now I get same results as you. prime95 can eror when everything else is stable. In my case it has only been a difference of a few fsb or one bump of .025vcore...