Hi I was wondering if anyone has OCed a P4 3.2, if so how far? Right now I'm oced @ 3.5, but due to the stock heatsink I can't really go much further until I get my watercooling setup going (about 2 weeks). Thanks for any input. I'm running with a Abit IC7-MAX3 1gb Corsair TwinX-4000Pro ATI Sapphire 9800Pro 128mb.

http://www.extremeoverclocking.com/reviews/cooling/Asetek_WaterChill_7.html Here is a review with water cooling, IC7-max and P4 3.2. Accordingly, you would only gain another 200mhz from where you are at. I think this is the range Intel P4 tops out.

By the way, P4s' do not have low idle temp despite some of the reports you read here. Check some of the reviews. I am running 5 P4's in my office, none idle below 37C at 70-75 F room temp. In fact, 40-42 is the normal idle. Since, I run a business, my expectation of stability is 99.99%. Stability is a very loose term; one crash in 10 hours can be acceptable to some or two crashes/lock ups twice a week is okay. And, I won't answer the question if I don't have the equipment.

Thanks for the help. Damn that review doesn't give me much hope. Oh well, maybe I'll get lucky. Also aren't those water temps a little high in the review?

If you read carefully they could get even higher because the test was done without case (on bench). I calculate the 3.2 clocked to 3.5 - 3.6 with 1.6 to 1.7 vcore would be around 105 - 110 watts. I have been using a special engineering prototype fan duct with Intel Stock heatsink for around 6 months. All 5 P4's are clocked from 40% to 62% over default. The max temp never exceeds 58C. So, there is a bit over 10C headroom remains. The negative effect is noise. The shape of the fan duct creates more wind noise.

It is very very hard to get a P4 above 3.6 on any system, in fact it is easier to get a P4 2.6 up to 3.4 than it is to get a 3.2 up to 3.6. With water cooling you may be able to hit the 3.8 to 4 GHz mark, but you'd have to be really careful with everything else, even if your system is cooled well enough your mobo and other components would begin to lose stability before you got much higher.

The wattage of an overvolted and/or overclocked cpu cpu is: stock wattage x (oc vcore/stock vcore) x (oc vcore/stock vcore) x (oc ghz / stock ghz ) because: Power = wattage = P, Current = I, Voltage = V, resistance = R P = I * R P = V^2 / R P= I^2 * R The resistance of a cpu is constant at constant frequency so wattage goes up a function of voltage squared if you increase the voltage since wattage = V^2 * R But even at constant voltage the current draw of a cpu changes proportional to frequency due to the charging capacitance of the chip, so since wattage = V * I, with voltage constant the wattage goes up proportional to current, which goes up proportional to frequency. So if both frequency (mhz) and vcore are changed, the increase in wattage is the formula in bold above, which this web page has buried in its wattage calculation field: http://www.benchtest.com/calc.html