Is a fan that is twice as expensive also twice as good?

Wed 18 Mar 2015, 00:00

Background

For months now, I have been cooling my CPU with a Cooler Master Seidon 120V. As the stock fan that came with it was horribly loud I removed it and replaced it with one of the fans that came with my computer case, a Fractal Design R3 (or maybe R2, not sure, I bought it pre-owned).

Anyway, the fan that I took was this one:

After googling it, it turns out to be a Fractal Design Silent Series 120mm, a fan that costs 99 SEK at Inet.se at the time of writing.

One day I ordered a new PSU from Inet.se (Fractal Design Edison M 650W), and I decided to buy a new 120mm fan then for my radiator. So I went for the Noctua NF-F12 PWM, which costs 199 SEK at Inet.se at the time of writing.

For fun, I decided to benchmark my CPU temps with this current fan, so I can compare it to the Noctua one to see if it made any difference.

The Computer

Gigabyte GA-Z97P-D3
Intel i7-4770K @ 3.5GHz (Overclocked to 4.1GHz @ 1.15 V)
2x4 GB Corsair Vengeance LP DDR3 1660MHz
ASUS R9 280 3GB DirectCUII TOP
Fractal Design Redefine R2/R3
Cooler Master Seidon 120V

I also have 2x 120mm fans in the front (in), and on at the top of the computer (out). The radiator is at the back of the computer with a fan blowing air from the inside of the computer outside (through the radiator).

The Tests

OCCT CPU test for 15 minutes
Prime95 Small FFTs until temp hits 100 C, or computer BSOD's

I ran the tests with my Fractal Design fan, and then I ran them again when I got the Noctua fan. While I ran the tests I had HWMonitor (to measure temps) and CPUZ running (to see CPU freq and voltage). I then used the freeware utility AutoScreenShot to take a screenshot of my screen every X seconds (3 secs for prime95, 15 secs for OCCT). After having run the tests, I checked the screenshots to get temperature, frequency and fan RPM. I wrote these down in an Excel document.

I did not take my room temperature into consideration during the tests. My room is usually around 21-22 C. This might cause some inaccuracies but I still think the results I got were because of the fans, not the room temperature.

Results

Conclusion

The Noctua fan beats the FD fan. It's not a huge difference, but it's an improvement. With the FD fan, Prime95 reached 100 C after 2 mins, while the Noctua fan kept going until 4:24 in, where the computer blue screened (I got a bad 4770k for overclocking I think, 4.1GHz @ 1.15 V should be easy), it never reached 100 C, it topped out on 97 C.

It is worth noting that for the Noctua test, my computer did use a different PSU. I don't think this made any difference to the temperatures though. If anything, it made it worse as prime95 has made my computer BSOD alot more since I got it.

So is a fan that is twice as expensive also twice as good? No, it isnt. It is better, but not by a huge margin.

If you want to, check out the excel document. It has all the data, along with fan RPM. The CPU never throttled down from 4.1GHz so those numbers are a bit uneccessary, but oh well. The Noctua fan had an PWM so that is why it increased it's speed.