To test the heatsink, we boot into Windows and let the system sit idle for 10 minutes, then measure the core temperatures. We then load the GPU core to 100% using FurMark until the temperatures level off and stabilize and then again record the core temperatures. GPU-Z was used to record all temperatures.

We tested the Prolimatech MK-13 on a Radeon 5870, comparing it to the reference design stock cooler, as well as a Sapphire Radeon 5870 Vapor-X card/cooler. Lastly, we tested the MK-13 in both the flush and perpendicular mounting positions.

So how does the MK-13 perform? Let's take a look below.

Interesting. A few observations and conclusions can be gleaned here. First, clearly the side mounting method is significantly better than the bottom mounting on the MK-13. The difference is an astounding 24°C with the high output Scythe fans and 18°C with the low noise Noctua fans. Further, when using the bottom mounting method, the VRMs run far hotter since the airflow cannot easily penetrate through the fins, with an increase of 21°C. And with the Noctua fans on the bottom, the VRMs hit 100°C.

We can also confirm that the side mounting method, even with the low noise Noctua fans, is not only cooler than the Sapphire Radeon 5870 Vapor-X but also quieter. In this configuration, the MK-13 is extremely impressive as a daily-use solution. Certainly the high output Scythe fans offer better temperatures but at much higher noise levels, far louder than the Vapor-X. And as far as the stock 5870 cooler goes, there really isn't any comparison, as even the "worst" configuration of the MK-13 still trounces it.

So what's the downside here in terms of the MK-13's performance? Well, the warm air isn't exhausted directly out of the case as is done with the Vapor-X, for example. In order to take advantage of the MK-13's impressive cooling abilities, you must have good airflow within your case to exhaust the warm air.