[cmeisenzahl]

Is this considered safe to use? Will it cause notable premature wear on the fans? Or is the jury still out?
smcFanControl2

Your thoughts?

Thanks in advance!
Chris

[frdmfghtr]

I've tried it, but never saw or felt any benefit. I've never had an issue with a warm enclosure bottom, and never saw any change in battery life. Some claim that the power savings in running a cooler CPU would offset the extra power sent to the fans, but I never saw that.

[Simon]

Originally Posted by frdmfghtr 

Some claim that the power savings in running a cooler CPU would offset the extra power sent to the fans, but I never saw that.

Those people are fooling themselves. If you crank up the fans, the CPU will become cooler because it's being cooled more efficiently, not because it's suddenly and magically using less power. At a given amount of load the CPU will use a certain amount of power regardless of the fan speed. The CPU temp OTOH is a result of the dissipated power and the cooling efficieny. The fact remains though: cranking up the fans will use more power - although that increase won't be dramatic compared to the usual power consumption of such a system.

Personally I don't use smcFanControl often because most of the time I don't mind a warm bottom case, but I do mind noise. There are however situations when you'd like to cool it down - for example when you're running full load and you have the MBP on your lap. In such cases it's perfectly fine to use smcFanControl.

[Simon]

dp, sorry

[mduell]

Simon: What you're forgetting is that semiconductors consume more power (at the same load) when their operating temperature increases. And it may be enough to outweigh the increased power consumption of the fan.

IBM used to have great detailed docs on their website about PPC970 power consumption; the difference in power consumption between 65C and 105C was ~5% IIRC.

[Simon]

Originally Posted by mduell 

Simon: What you're forgetting is that semiconductors consume more power (at the same load) when their operating temperature increases.

I'm actually not forgetting it - as a research scientist in a solid state physics lab I shouldn't either!

The reasoning is that we know from our own experiments but also from literature that the power draw increase due to moderate junction temperature increase is extremely small. Of course ohmic resistance and transistor leakage are always functions of temperature, but in the case of CPUs like the C2D this translates to a very weak dependence in the consumed power. You can heat a C2D from 40C to 80C and you'll observe a single digit percent increase in power draw at best (also illustrated by your PPC970 example even though the 970 is actually a bad example since power density at its 'hot spots' is extremely high). Compared to Merom's TDP of 35 W and the power consumption of other components the sub-W draw increase due to an additional 20C become negligible. And of course it becomes even less if you factor in that the fan you just cranked up to get that cooling also draws up to 2.5 W. So yes, of course power draw depends on junction temperature, but does a 10C cooler CPU give you more battery life? No.

[drnkn_stylz]

Wow, Simon, that is some solid info right there. Good thread guys!

[Randman]

I use it for intensive gaming (WoW) but set it back after that.

[The Placid Casual]

Hmm not seen this one before, was out of the loop for a while!

My machine gets very hot playing WoW sometimes, so might crank the fans up while playing...