Quote:
Originally posted by dhamsaic
Secondly, I just placed the orders for the HSF (HeatSink Fan). I got two (2) ThermalRight SK6's with two (2) 60mm 38CFM (Cubic Feet/Minute) fans. The ThermalRight SK6's are all copper and have 70 fins each, which basically leads to superb cooling.
|
There is only one specification that is appropriate to heatsinks. Degree C per watt. Number of fins, material, CFM on the fan, all this is irrelevant since only the bottom line applies. If they make a decent heatsink, the first number provided is Degree C per watt (and not CFM).
How to measure CPU fan/heatsink quality. Any properly machined heatsink can be applied to the CPU without thermal compound. The experiment is quite simple and decades old. Attach the heatsink without thermal compound, test the system using standard programs, and measure the CPU temperature. Now repeat the same test with thermal compound applied. If CPU temperature drops by more than single digit degrees, then the heatsink was machined inferior.
Despite all the hype with more Cubic Feet per Minute of air and high tech thermal compound, little has changed in 20+ years. Thermal compounds have the same thermal resistances of past decades. Heatsinks have only gotten bigger meaning that 1950 machine processes are no longer acceptable for current heatsink manufacture. For all the hype associated with heatsink fans, why is the Degree per watt number so hard to obtain? It is the first number an engineer uses for his thermal design.
BTW, any engineered system into the 700 Mhz range could cool just fine without heatsink fans. Passive cooling is preferred in more reliable systems of that time. Today's 1 ghz CPUs can be cooled with passive heatsinks, but there is less room for error. Among factors so critical to passive cooling designs is Degree per watt (CFM of the fan is a useless number). So why is that number so difficult to get from so many hyped CPU heatsink manufacturers? Maybe the above simple experiment could provide answers.