Cooling Efficiency and Acoustic Performance

First of all let’s check how well Zalman VF2000 LED copes with a graphics card. However, I have to say right away that unfortunately, we couldn’t complete the graphics card test with Zalman VF2000 LED cooler in FurMark for the following reason:

The GPU temperature was unusually slow even for a hot graphics card like that, but the VRM temperature reached152 °C 42 seconds into the test. That is why we had to terminate the test at this point to avoid a fatal outcome for our quite expensive graphics accelerator. However, there is nothing unexpected here, because leaving the voltage regulator components of a Radeon HD 4870 without any cooling (as you remember, we are testing Zalman VF2000 LED as is) is practically the same as accelerating on a Formula 1 with a covered engine air duct.

As for the reference cooler of Radeon HD 4870, it did do its job in all test modes, even though the noise level was pretty high:

You can see from the screenshots above that the graphics processor temperature with the reference cooler in play is higher than that with the Zalman VF2000 LED. The same is true according to the results of 3DMark 2006 test, where the voltage regulator circuitry is loaded less and where the new cooler worked perfectly fine:

In terms of VRM temperature, Zalman VF2000 LED loses again to the reference cooler of the Radeon HD 4870, but this is something we have actually expected. However, the fact that the new Zalman solution easily defeated the massive copper Radeon cooler in GPU temperatures turned out quite a surprise. I have to admit that I didn’t expect a relatively modest and small cooler like Zalman VF2000 LED to demonstrate such remarkable efficiency. Note that Zalman’s advantage over the reference cooler often reaches up to 20 °C! if you remember, not too long ago when we tested a three times larger Arctic Cooling Accelero Extreme GTX cooler it also showed a remarkable advantage over the GeForce GTX 260 reference cooler. And most interestingly, it also had the same problems with VRM cooling! Knowing about it Zalman took care of the issue having released a special ZM-RHS90 unit intended for the cooling of these particular components shortly after launching their Zalman VF2000 LED.

Now let’s take a look at the cooling efficiency of our testing participants during CPU tests:

I have to say that it showed tremendous cooling potential. I doubt that anyone hoped to see the low-profile Zalman VF2000 LED as a serious rival to one of the most efficient top-coolers. Simply because it wasn’t the manufacturer’s ultimate goal. The mere fact that it is possible to overclock a processor with this cooling solution on it is a great achievement already, especially since it can be done at the minimal fan rotation speed. At maximum fan rotation speed AMD Phenom II X4 remained stable up until 3.65 GHz without any additional voltage increase, which is a very good result for a 260 g cooler as tall as only 45 mm.

As for the level of noise the new solution generates, everything is pretty simple. AT minimal fan rotation speed of 1350 RPM you can’t hear Zalman VF2000 LED against the background of a working system case with quiet components inside (less than 33 dBA at a 1 m distance). The cooler remains within acoustic comfort zone up until 1850 RPM when its noise exceeds 34.5 dBA. And after that in the interval from 1850 to 2450 RPM the cooler noise gradually increases up until 38.9 dBA and I doubt that anyone will be using it constantly in this loud mode. Here Zalman VF2000 LED yields to Scythe Kabuto quite significantly, but the reference Radeon HD 4870 cooler again remains an outcast, because its noise level is tremendous. We didn’t notice any parasitic noises coming from the Zalman VF2000 LED fan, such as crackling, howling and etc. So, now it is time to sum everything up.