[KLUG Hardware] Re: Heat Sink compound -- exposed FC-PGA v. w/heat spreader

[randall perry]

At 04:58 PM 1/17/2003 -0500, you wrote:


>Quoting randall perry <randallp@domain-logic.com>:
>> You gotta watch it with the large PPro or older Xeon (huge processors). 
>> With such a large area you have to use a thermal pad or mesh-reinforced
>> paste.
>
>???
This isn't where I first read about it, but another reference to it:
http://www.arcticsilver.com/as3.htm


>> Yeah, the core is too tiny to pump that much heat away from the
>> processor
>
>Er, it's a give/take relationship.  The reason why you want to make sure you
>have grease on the outside of an exposed flip-chip is to _increase_ the total
>surface area of any transfer, because the area is so much smaller than one with
>a heat spreader atop (like newer P4s).
>
>But an exposed flip-chip _without_ a heat spreader actually _increases_ thermal
>transfer by an average of 2-3 degrees C.  Why?  Because there is _nothing_
>between the die and the heatsink, other than an extremely thin plastic covering.
>
>Several enthusiast sites have backed this up with testing of P4 chips without a
>heat spreader, as well as AMD chips with an added "copper cap."  Which is why
>AMD is continuing to release FC-PGA processors without a heat spreader, unlike
>Intel.
>
>> Yeah, the Artic is good shtuff.
>> They don't use silicon in their compound and they actually screen more
>> than one size of ceramic dust through so that you get better arranged
>> particles (and better heat transfer).  It also cleans up pretty easy. 
>> Not like that crap dot of compound you get with your heatsink (or
>> thermaltape)
>
>Correct.