[luau] Underclocking / Aluminum cases

[Joe Linux]

But the issue is still how to do it without a fan motor.  In the example 
of the Dell, they are substituting a case fan for a fan on the processor 
and it would have to have the case on to work properly.  It's well known 
that ducting and cowlings on fans improve efficiency.  The fact that you 
do a lot of little things to improve air flow shows you are both 
thoughtful and clever.

MonMotha wrote:

> I guess you've never used a Dell Dimension XPS Dxxx series tower 
> system.  The processor has a big heatsink but no fan on it.  Directly 
> behind the processor is the case fan.  Take the case off an run the 
> CPU at 100% for long, proc overheats and halts.  With the case on, run 
> at 100% CPU for days, the heatsink is barely warm to touch.
>
> I've actually built white boxes with similar, though not so drastic, 
> thermal characteristics.  When I build systems, I usually consider 
> airflow from the start, even down to selecting different brands of 
> video cards due to how far back they extend!  I route cables behind 
> things, twist tie cables next to bracing, etc. all to improve airflow 
> from front to back (I've even built a wacky system where I flipped the 
> PSU fan and had air flowing from back to front).  Most of my systems, 
> when I'm done,  have case temps within 0-3C of room temp, and CPU 
> temps within 0-10C of case temp, but I've had systems where, due to 
> good airflow, the CPU runs hotter with the case off (even though 
> sometimes ambient system temp is lower) than with the case on.  Good 
> airflow allows for faster evacuation of the hot air produced by the 
> CPU fan circulating air over the heatsink, thereby lowering the 
> ambient temperature NEAR THE CPU, creating a greater delta T between 
> the heatsink and air, allowing for faster heat transfer from the CPU 
> heatsink to the air.
>
> --MonMotha
>