[Ti] Macbook pro version 2...

[Kynan Shook]

Chris Olson <chris.olson at astcomm.net> writes:
> Your data is inaccurate.  Show me what production Intel cpu you've
> ever seen that spec'd out agreeing with the product data sheet?

And I'm supposed to believe you on this?  The data I used was from  
the recommended maximum current supply.  If they're exceeding that,  
then you'd have designs that won't work because there's too much  
voltage drop on the PCB's Vdd planes.

> You can grab a 40 watt light bulb and it's barely warn to the touch.

You must have hands made of asbestos.

> Any cpu that gets that hot simply because of misapplied thermal
> conductivity paste pulls more than 31 watts.  Like I said, Intel
> tends to be horribly optimistic.  For pete's sake, you can cool the
> hottest-running G4's with a passive heat sink that takes up way less
> room than the thermal pipes and sink on the Core Duo.  Anybody with a
> brain should be able to read between the lines.

Again, you're talking out of some orifice that I have to assume is  
not your mouth.  I'm currently designing a board whose main component  
dissipates less than 8 watts.  It has a medium-sized heatsink, uses a  
flip-chip package (aka there is very little thermal resistance  
between the silicon and the heatsink) and our thermal simulations  
show that we will need at least 150 linear feet per minute of airflow  
to keep it cool under the full temperature range.  Our operating  
temperature goes higher than Apple's (55°C instead of 35°C), but so  
does the maximum junction temperature (125°C instead of 100°C).

So if the thermal paste is in fact causing problems, it's not  
terribly unreasonable, because any miniscule changes in the cooling  
system can have a huge effect.  And I haven't seen the cooling system  
in the MacBook Pro, however I've seen the one from the 17" PowerBook  
G4, and it's rather large too.  Seeing as how there's not much room  
for it to increase in size, I have to assume it's similar in the  
MacBook Pro.