Tuniq Tower 120 Cooler Review

PERFORMANCE


All tests were made on my main system:

• Antec LANBoy case with one 120mm fan in the front and 1 120mm fan in the back.
• AMD64 3200+
• ASUS nForce4 K8N…. SLI motherboard
• 1 GB OCZ PC3200 RAM
• Sunbeam NUOO PSU
• One 36 GB 10000 rpm Western Digital Raptor HD 
• One 200 GB 7200 rpm Maxtor DiamondMax 10 HD
• Stock cooler for the AMD64 3200+
• Tuniq Tower 120

The CPU temp and fan rpm was measured with ASUS PC Probe. It reported slightly higher rpm’s than Tuniq says the fan is capable of (2136 rpm as max and 1222 rpm as min compared to the 2000 and 1000 rpm in the documentation).

To test the coolers we put them through a few tests designed to work the CPU:

• The CPU-tests in SiSoft Sandra Pro 2005
• The CPU-test in 3DMark2005

These were run in loops and the max temperature was noted. In addition to running the CPU at default speed (200x10) we also overclocked it to 220x10. We then did run the same tests and noted the max temperature.

As you can see the stock fan cools the CPU a bit worse than the Tower 120. It isn’t until we overclock the CPU to 2.2 GHz (a 10% overclock) and turn down the fanspeed to the minimum that we see similar temperatures with the Tower 120. If I had tried the stock cooler at 2.2 GHz I’m pretty sure we would have seen higher temperatures.

Noise

So we’ve seen that the Tower 120 can cool the CPU better than the stock cooler. However, what about noise? The stock cooler isn’t the noisiest fan I’ve hard but it most certainly isn’t silent. The Tower 120 actually has 2 faces. At the max RPM (~2000) it is noisy. I’m not sure I would like to run it at that speed all the time. At the minimum speed (~1000) things are different. Now the fan is very silent and the other fans in the system are louder. The fact that it still cools very well at the lowest speed makes it possible to run the machine at that speed most of the time while you can turn it up temporarily if you really need the extra cooling.

From the chart it looks like the noise picks up around 1800 rpm. I could verify this by going up to around 1800 rpm before noticing a larger increase in the noiselevel. There’s a lot of room for trial and error to find the best compromise between cooling and a low noiselevel.