Build Quality
The build quality of the NH-L12 package is up to Noctua’s usual standards. The solid mounting system as seen with the NH-D14 has been implemented here and combined with a pair of quality fans. The cable sleeving was impeccable and the inclusion of Noctua standard accessories included:
- extension cables
- splitters
- voltage reducers
- mounting brackets
- thermal compound
Testing
The following tests were run in our Fractal Design Define R3 with the side panel open so we could hear the noise made by the NH-L12. The fans were configured to run off the ASUS Standard PWM profile – where indicated, we installed the voltage reducers to cap the top speed of the fans. We recorded the delta temperatures – the difference between the reported CPU temperature and the ambient room temperature. In order to record a valid result, the ambient room temperature has to remain the same from the start of the test to the end of the test and must be between 18 and 24 degrees.
A Note from the Manufacturer:
Noctua have the following statement on their website relating to this product:
“Caution: The NH-L12 is a low-profile quiet cooler designed for use in small form factor cases and HTPC environments. While it provides first rate performance in its class, it is not suitable for overclocking and should be used with care on CPUs with more than 95W TDP (Thermal Design Power). Please consult our TDP guidelines to find out whether the NH-L12 is recommended for your CPU.”
So we read the above and thought that our 2500K (95W TDP) CPU would be a good test for a worst case scenario. Overclocking it would be pushing the boundaries beyond reason …. so we did that too.
Test Rig
- Intel i5-2500K Sandy Bridge CPU
- ASUS P8Z68-V Pro Gen3 Motherboard
- 16 GB of Corsair Vengeance DDR3 Blue Low Profile RAM
- Fractal Design Define R3 Case – White with USB 3
- Intel Series 510 SSD – 120GB
- Corsair HX-650 Power Supply
- 7x 2TB SATA Hard Drives
With Both Fans:
In Ultra Low Profile mode with only the 90mm fan fitted:
Fractal Design Define R3 Build log can be found: HERE
Delta Temperature Results
| Scenario | CPU Settings | Fan Speed | NH-D14 | NH-L12 | NH-L12 – NF-B9 Body |
| Idle | Stock (no OC) | 12 Volts | 8.5 | 7.5 | 9.5 |
| Load | Stock (no OC) | 12 Volts | 31.75 | 44.5 | 58 |
| Idle | Stock (no OC) | L.N.A | 9.25 | 9.0 | 9 |
| Load | Stock (no OC) | L.N.A | 33.75 | 47.0 | 64 |
| Idle | OC @ 4.5 / 1.288v | 12 Volts | 7.25 | 6.5 | – |
| Load | OC @ 4.5 / 1.288v | 12 Volts | 34.5 | 53.0 | – |
| Idle | OC @ 4.5 / 1.288v | L.N.A | 9 | 7.0 | – |
| Load | OC @ 4.5 / 1.288v | L.N.A | 36.25 | 55.3 | – |
| Idle | Stock (no OC) | PWM | – | 8.5 | 9.5 |
| Load | Stock (no OC) | PWM | – | 42.3 | 54.75 |
| Idle | OC @ 4.5 / 1.288v | PWM | – | 7.5 | – |
| Load | OC @ 4.5 / 1.288v | PWM | – | 55.3 | – |
Notes:
1. The NH-D14 does not have PWM capability and as such, there are no results for that product in PWM mode.
2. The NH-L12 when operating with only the 90 mm NF-B9 fan was not able to adequately cool the 2500K when overclocked at 1.288v and belted with Prime95 so these tests were omitted as the results couldn’t be finalised.
Idle Delta Temperatures
The Idle temperatures are all within a typical margin of error. There is only so low the idle temperature can go an picking idle temps can be finicky because they can change in the blink of an eye. We have included them in the interests of being thorough as the idle temperatures are relevant but for a real comparison, we will be looking at and talking about the load Delta temperatures.
Load Delta Temperatures
When running the NH-L12 in single fan mode, the results with an i5-2500K at stock settings were understandably higher but still well within the boundaries of safe.
When used in single fan mode on an overclocked 2500K, the NH-L12 could not keep the CPU temperature below 90 degrees during the test so the result was “technically” a fail. This was expected due to the product being used in a way that Noctua specifically recommend against – but we tried it anyway, monitored our temps closely and aborted the runs accordingly.
