Anathema Wrote the following on 10/21/2005 6:30 PM : This video is a good example of Toms Hardware selling out. They try and make the intel chips look better by saying an intel will survive a catastrophic heatsink failure and an AMD wont, nevemind the fact that this sort of failure is incredibly rare. Did you see how much trouble he had getting the last heatsink off? that was the only one he actually attached to the motherboard. I cannot think of anyone who has ever had a heatsink fall off their processor.
In the real world, AMD chips endure much higher operating tempatures. it's not unusual for Barton core Athlon XP's to run at 65C under load, a tempature that would cause serious problems for a p4.
Athlons handle Real world overheating conditions much better then p4's, and their smaller pipeline and other architectural differences allow it to process information far more effectively, at lower clock speeds then the p4.
Intels day in the sun is over, They have reached the practical limits of the p4 architecture, while AMD is just getting started. They got lazy because of their big deals with Dell and other manufactuers.
That being said, It's a very cool video to watch. I've smoked a processor or 2 in my life.
Surfer Wrote the following on 10/21/2005 6:50 PM : The concept is to simulate a cooling fan *failure*, not one falling off. And while it is rare as far as hardware failures go, it's not incredibly unheard of. And no, 65 C is no problem for a P4 (I know this for a fact, I run the hell out of mine and get it up to 68 *regularly* with no problem at all, and that's with it overclocked - I built my current PC for high performance and with possible heat problems in mind.)
As far as the usability of Pentium vs. AMD, pentiums can be milked for longer periods of time by tweaking and targeted software applications and optimizations. I've had a 500 P3 run circles around a 1.1 Ghz AMD in every aspect (I benchmarked both of them).
But since the point is brought up, perhaps there's been some studies done that show AMD's surviving a failure like that?
Anathema Wrote the following on 10/22/2005 04:00 AM : If you wanted to simulate a cooling fan failure, you'd unplug the processor fan, not remove the heatsink.
Surfer Wrote the following on 10/22/2005 09:40 AM : You *could* do that, but then how the hell would you take the temperature of the processor? Jesus, quit being a damned idiot and think for yourself! They followed the simple and time proven process of hypothesis, prediction, experiment, observation (that process followed in science) and you want to bitch about it.