p-3 550e @825

[newkilla]

hey guys
talk about weird,i posted 2 messages
in 3 forums for help with the p3--i got zilch
got to give credit where credit is do.
enter the mad man aka bruno,without whose help id still be laughing like ***** wood pecker.lol,i dont understand this system at all.pc 100 ram at 150 fsb with default voltage and stock cooling. go figure????????
thanks again
new---killa

[akaBruno]

Thanks Dave. All it really took was a boot to your ****, to get ya jumpin' that thing. The other problem was the fact that nobody's familiar w/ that Asus P3V 133 board. But, she's Shmokin now !!! Download that new driver for it.

Bruno

[krusty the klown]

Newkilla, you can run PC100 mem and a 150MHz FSB - you just can't run the mem @ 150MHz (Errr... unlikely, anyway).

Not sure what the BX boards let you do, but VIA 133 boards let you run mem at FSB-PCI bus speed, so you can run the PC100 @ 150-37.5 = 112.5MHz. They also let you choose a /2 (/4) dividers to get reasonable AGP (PCI) clock speeds when running such high FSB's.

You can also run the Coppermine P3's at such high speeds with relatively low cooling requirements. These chips are the 0.18micron fabrication process, which allows Inhell to run much lower core voltages (1.65V) and so the chips produce less heat.

If you have a 550E which can do 825 @ default voltage, it's a good chip - mine will do this with a 7.5% increase in core voltage. Now the weather's finally getting warmer in the UK, my temps are starting to rise, so I'm investing in a VOS32 cooler (a beast) - see:
http://sysopt.earthweb.com/forum/Forum3/HTML/006481.html

for a piccie http://sysopt.earthweb.com/forum/smile.gif

Hope this clears a few things up http://sysopt.earthweb.com/forum/smile.gif

[newkilla]

sounds good to me, i understand if i want to go the gforce route im going to have to play with the voltage a little whether it b the i/o or core , i cant find it anywhere there is a jumper for both,the selection being
default and test the manual says test might b beneficial in overclocking applications
but they strongly advise not to for you will lower the life of the chip,blah blah blah
you"D think we were talking about a puppy or something.NO ONE LIVES FOREVA any help would b great
newkilla

[akaBruno]

OOOOOhhhhh, so the setting is 150mhz(fsb)minus 37.5mhz(RAM)= 112.4mhz(RAM). That's why there's a second set of jumpers for the divider.
I think I get it now. Hmmmmm...
Thanks Krusty,
Bruno

[krusty the klown]

You got it AKABruno!

If you're mad enough, there's also the option (certainly on my mobo - Soyo SY-6VBA-133) to run the RAM @ FSB + PCI bus!!!! Hehe, I'd like to see it try and clock @ 187.5!

I guess this is only useful if, for example you had a celery running at 100MHz FSB (overclocked) and you invested in PC133 RAM. You could then clock your RAM @ 133, though I'm not sure what performance increase this would offer (maybe best to run @ 100 and CAS2 / 'fast' settings, in this case).

Newkilla, some mobos appear to have a facility to increase the i/o voltage for AGP (and PCI??) sockets (I've heard this mentioned a couple of times). I guess this can be helpful if you're clocking your AGP bus way over spec (like on BX boards with > 133MHz FSB). My mobo doesn't have this facility, so I have no personal experience of this, but it will let you increase the CPU core voltage by +2.5, +5, +7.5 and +10%.

I think you're real enemy when increasing voltages is heat. I think that's the most likely culprit for killing chips early. For example, you can kill a non-OCed chip running @ default core voltage if the fan fails.

Bear in mind, that if you increase the core voltage by a certain %age, the power dissipated increases by more than this:

Power = ((volts)^2)/resistance)

if we say the resistance = 1, for simplicity and we have a 2.0V chip.

If I increase the voltage by 10%, the power increases by 21%

power @ 2.0V = 2^2 = 4.0
power @ 2.2V = 2.2^2 = 4.84

4.84/4.0 = 1.21, a 21% increase.

So be prepared to have to sort out better cooling.

I was going to post a thread asking whether a chip running @ increased core volts, but with sufficient cooling so that it ran the same temp as a default voltage chip (with a smaller HS) would die quicker. Now you've sort of brought it up, I will http://sysopt.earthweb.com/forum/smile.gif