Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Not anymore...

If you are in the mood for reading further well researched articles based on facts, go have a look at Anandtech. They did a series in early 2006 when Apple made the switch to Intel.

Even Apple touted performance per watt in the keynotes around the Intel switch. They'll be on YouTube.
 
Back on topic...

I've been stress-testing the CPUs since repasting last night. Neither has gone over 130ºF, while "A" was pretty reliably running 136-137ºF under load before and "B" about 126-127ºF. Now A is running at 128-129º, and B 124-125º.

That's not a huge drop in temperature, but I'd call myself satisfied.
 
Back on topic...

I've been stress-testing the CPUs since repasting last night. Neither has gone over 130ºF, while "A" was pretty reliably running 136-137ºF under load before and "B" about 126-127ºF. Now A is running at 128-129º, and B 124-125º.

That's not a huge drop in temperature, but I'd call myself satisfied.

Temps should improve a bit more over time as the thermal paste sets up.

I should do the same to my xserve g4, its paste is 12 years old. Temps on it are very stable at 55c under heavy load.
 
Temps should improve a bit more over time as the thermal paste sets up.

I should do the same to my xserve g4, its paste is 12 years old. Temps on it are very stable at 55c under heavy load.

I know that AS5(which is what I use) says to allow 200 hours of "cure time" with hot/cold cycles in between. That's part of why I'm stress testing it with the CPUs pegged for a few hours, and then allowing it to cool down. I probably should throw in a complete power-off for a few hours somewhere in there too.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.