As a general rule, we prefer to test cooling devices on a synthetic test bed, but sometimes that just isn't possible. To test out the Zalman ZM80- HP we've enlisted the help of an Albatron GF4 ti4600 videocard since it is the type of card most likely to be used with the ZM80-HPVGA cooler.We have opted to place the thermocouple from the Omega HH501DK thermometer directly behind the GPU in each instance for temperature measurements.

This might seem a little odd at first as the instinct of most people would be to place the temperature sensor directly above the requisite GPU core, or just adjacent to it. After factoring in the option of drilling tiny little holes for the type-K thermocouple everywhere, and the multitude of stock heatsinks we would be contending with, not to mention differing types of BGA packages, simplicity reigned supreme.

Throughout this series of reviews on videocard BGA coolers, we will be taking temperature measurements with a Type-K thermocouple placed directly behind the core, effectively measuring the temperature of that small patch of PCB.

The thermocouple is insulated against the surrounding environment with a small section of closed-cell foam. The foam helps to ensure that the thermocouple stays put, and isn't cooled off by exhaust air eminating from the processors' heatsink, case fans, or any other transient sources.

To generate some heat in the actual GPU, we found that running 3DMark2001 three or four times in a row nets fairly consistent results, though different GPU's react differently to the tests which is quite interesting.

We've also included a set of results with a 4000RPM 92mm case fan blowing in the direction of the Zalman ZM80-HP (simulating a front case fan which is common on most computer enclosures) from five inches away. This fan does create a small amount of noise, but as the ZM80-HP does not ship with any fans, the results are mainly intended as an example of what active cooling could bring to the table.

	FrostyTech VGA Card Heatsink Results
	Manufacturer	Heatsink Model	Avg. Ambient Temp.	Nvidia Ti4600 Temp.
	Zalman	ZM80-HP	23.1	66.3	Passive Cooling Only
	Zalman	ZM80-HP	23.1	52.6	With Active Cooling
		Stock Active HSF	23.1	54.8

Looking at the temperature results there are two items to take note of right away. First off - the temperature of the card in fully silent operation will rise a few degrees (about 10C)above that of the actively cooled stock heatsink. This is expected, and while we would obviously like to have seen the temperature go the other way, it really only makes sense.

The Zalman ZM80-HP is intended for silently cooling stock videocards, and in that respect it works. Temperatures will be higher due to the passive nature of the cooling system, but the videocard remains completely silent. If we take a few liberties with the Zalman ZM80-HP and add a low-noise 92mm case fan a few inches away, but pointing at the cooler the tables turn. In that second case, the ZM80-HP is able to drop the temperature of the card just a little below that of the stock actively cooled heatsink.

The bottom line is that the ZM80-HP works well for nVidia GeForce 4 cards as we have shown here, with a raised temperature under passive silent cooling, and with a lower temperature under a small breeze from case fans. In either case, the Zalman ZM80-HP is an incredibly innovative product for videocard cooling and an excellent solution to the noise created by high-end videocards.