Is it worth it?

[ToolBox]

Maxvla [I hope thats right], the senior member, said that on a P4 watercooling isn't much better than good air cooling. Is this true? Is it worth it to watercool on a P4?

 

[alpha_gamer]

Just go for a really good cooler like the SLK 900 from Thermalright, unless you're doing some insane overclocking this will work very well.

 

[johan851]

I'd say that's about right. A good heatsink should do fine. One reason is that the P4's won't get as hot as AMD's when overclocked. AMDs tend to need a big voltage push to get their max potential, while P4's usually don't get much performance gain once you increase vCore past 1.65v or so.

 

[repilce]

i agree, unless you are OC'n majorly, want a quiet cooler, or just a modd phreak then a good air cooler will probably do you just fine, but to WC is just so fun

 

[Cathar]

One reason is that the P4's won't get as hot as AMD's when overclocked.

Doesn't help that the P4 thermal probe doesn't actually report the CPU temperature.

P4's don't run cooler than AMD's. They just report lower because of how the thermal probe is implemented.

 

[ILikeMy240sx]

I think the major reason for watercooling a P4 is silence. Because of low voltage, good air-cooling will give you enough cooling capabilities.

 

[alpha_gamer]

The Bartons run pretty cool too, I'm not sure how they compare to P4's, but they run a lot colder than previous Athlons.

 

[johan851]

P4's don't run cooler than AMD's. They just report lower because of how the thermal probe is implemented.

Haven't heard that, but I'd believe it. But what I'm saying is the typical overclocked P4 at 1.65 volts is probably cooler than the typical overclocked AMD at 1.95v, simply because of the voltage. Is that correct?

 

[Wicked Klown]

With alot of the new AMDs you don't need alot of voltage to get a nice OC. With my Barton 2500 stock FSB of 166 I can run at 200FSB at stock Vcore but just to be safe I run at 1.70Vcore.

 

[FizzledFiend]

With alot of the new AMDs you don't need alot of voltage to get a nice OC. With my Barton 2500 stock FSB of 166 I can run at 200FSB at stock Vcore but just to be safe I run at 1.70Vcore.

what you saying m8? don't understand a word there!

 

[Cathar]

Haven't heard that, but I'd believe it. But what I'm saying is the typical overclocked P4 at 1.65 volts is probably cooler than the typical overclocked AMD at 1.95v, simply because of the voltage. Is that correct?

No, not really.

Intel are also a little sneaky. They report "typical thermal output" while AMD reports "peak thermal output". Intel CPU's "peak thermal output" is about 15% higher than the "typical" value.

If one runs through all the thermal physics of it, including the effects of the IHS, then there is absolutely no way that P4 CPU's run cooler than AMD CPU's.

That everything thinks that P4's run cooler is purely because the thermal probe is what they're basing that assumption on.

Better yet, we can use Intel's own words:

Read sections 6.3, and 6.3.1 of the following document:

ftp://download.intel.com/design/Pentium4/datashts/29864311.pdf

Basically the temperature reported by the thermal probe cannot be used to determine the behavior of the thermal monitor sensor that controls thermal throttling and thermal shutdown.

i.e. what the thermal probe reports is not what the actual CPU temperature is. A second thermal probe measures that and controls the thermal functions of the CPU. It is important to note that this sensor is not visible to the motherboard.

A user-visible thermal sensor can, and often does, dramatically under-report what the actual CPU temperature is. There have been test cases created which demonstrate it under-reporting by up to 40C (!!).

 

[CrashOveride]

Boy O, you really do learn somthing everyday! Thnks Cathar.

But I would probably agree that for P4's top of the line Air is pretty close to a good watercooling system, but louder. But then again I am an AMDer so this is just what I seem to see most often when I am passing through Intel threads.

 

[johan851]

Fascinating information Cathar, I had no idea...

From the .pdf document:

The TCC reduces processor power consumption by modulating (starting and stopping) the internal processor core clocks. The temperature at which Thermal Monitor activates the thermal control circuit is not user configurable and is not software visible.

With a properly designed and characterized thermal solution, it is anticipated that the TCC would only be activated for very short periods of time when running the most power intensive applications. The processor performance impact due to these brief periods of TCC activation is expected to be so minor that it would be immeasurable. An under-designed thermal solution that is not able to prevent excessive activation of the TCC in the anticipated ambient environment may cause a noticeable performance loss, and in some cases may result in a TC that exceeds the specified maximum temperature and may affect the long-term reliability of the processor.

Basically, when the processor gets too hot (the "Trip Point") the TCC turns off some of the clocks to cool it back down. Ironically, this would mean that only when you're going for max performance (Games, benchmarks, high end graphics, etc.) might you experience a performance hit. I assume that it's not significant, but with a bad cooling solution you might not be getting good results with that P4.

I never knew anything like that Cathar, thanks for digging that up for me. Guess watercooling would be worth it for a P4...I stand corrected.