Good shot except that the fan gets a *lower* pitch when I turn seti *off*.
Seti is not the problem of course. I did try a simple C program as well:
main(){while(1);}
and the same thing happens.
I guess I'm still open to an explanation. :-)
--
-Time flies like the wind. Fruit flies like a banana. Stranger things have -
-happened but none stranger than this. Does your driver's license say Organ
-Donor?Black holes are where God divided by zero. Listen to me! We are all-
-individuals! What if this weren't a hypothetical question? [EMAIL PROTECTED]
On Sun, 18 Feb 2001, Bill Freeman wrote:
=>Steven W. Orr writes that a fan in his PC changes pitch depending on whether
=>he is running setiathome:
=>
=> My guess is that the fan runs slower because, without
=>setiathome running, the cpu really does run cooler, because it is
=>halted most of the time. Linux doesn't just burn cycles when it is
=>really idle, but halts the cpu. Interrupts from the clock tick and
=>other devices, like the keyboard and maybe network interface, wake it
=>up as necessary, but, assuming that it's a really fast cpu, that keeps
=>it un-halted only a small percentage of the time. Setiathome is
=>compute bound, so it runs instead of waiting on something (like the
=>screen saver effectively waiting on a clock tick), and the cpu never
=>halts.
=>
=> Any compute bound task should do the same thing. You can check
=>by writing a trivial delay loop in your favorite language (a real old
=>fashioned delay loop, no calling sleep). Run it with setiathome stopped
=>and the fan, I bet, will stay fast.
=>
=> The point of setiathome is that your cpu usually is doing
=>nothing. When the cpu is *really* doing nothing, Linux lets you save
=>power and heat (and some of the aging that heat causes).
=>
=> Bill
=>
**********************************************************
To unsubscribe from this list, send mail to
[EMAIL PROTECTED] with the following text in the
*body* (*not* the subject line) of the letter:
unsubscribe gnhlug
**********************************************************
