Page 3 - Test Results
Our test configurations are as follows:
CPU: Intel Core i5-4690K (Stock settings)
Motherboard: Gigabyte G1.Sniper Z87
RAM: Patriot Viper 3 Series Black Mamba DDR3 16GB (2x8GB)
Graphics: Gigabyte Radeon HD 7870 2GB
Chassis: SilverStone Kublai KL05B-W
Power: SilverStone Strider Gold S ST85F-GS 850W
Storage: SanDisk Ultra II 240GB; Western Digital Blue WD5000AAKS 500GB
Operating System: Microsoft Windows 7 Professional x64
Compared Hardware:
- Scythe Ninja 4
- Cooler Master Nepton 240M
- Noctua NH-L9x65 (Single/LNA)
- Noctua NH-L9x65 (Single)
- Noctua NH-C14 (Dual Fan)
- Noctua NH-C14 (Single Fan Bottom)
- Noctua NH-C14S (Top)
- Noctua NH-C14S (Bottom)
- Noctua NH-D15S
- Noctua NH-D15S (LNA)
- SilverStone Tundra TD02-E
- SilverStone Tundra TD02-LITE
- SilverStone Tundra TD03-E
- SilverStone Tundra TD03-LITE
- Intel Stock
All tests were run in our custom built computer to best reflect real life performance. The computer remained in the same place and room throughout all tests. The ambient room temperature in the room was around 22c. The thermal paste applied to each cooler was stock respective to their manufacturers to rate its performance; all pastes had sufficient time for them to fully settle. The fans on all heatsinks were connected to the same motherboard 4-pin connector. The test computer was turned on and idling for at least one hour for the idling tests. High CPU load results were obtained using Prime95 in-place large FFTs test with four worker threads for a minimum of ten minutes, and recorded when the temperature is deemed stable. The Intel stock cooler was chosen as a baseline reference. It is a good heatsink to figure out the delta between the reference unit and the tested product for standardizing performance against other coolers, even if not compared directly against. As you can see there are quite a few other CPU coolers that we are comparing the Scythe Ninja 4 to. Temperature results were measured with RealTemp, which reports the CPU's integrated digital thermal sensor for maximum accuracy. Each temperature result was calculated by taking the maximum value of the cores inside the CPU.
After idling the computer for the required time, the Scythe Ninja 4 reached a stable temperature of 28 degrees Celsius. This test was done with only one fan; it is possible for another fan to be mounted to have even better performance. Comparing this to some of the other tests, you can see that the Scythe Ninja 4 posted a good score for idling temperatures. Another thing to note is when the computer is idling, the fan on the heatsink slows down to around 300 RPM, making the computer really quiet. However, what we really want to see is how it did under load.
After using Prime95 to put a lot of stress on the CPU, the Scythe Ninja 4 stabilized at a respectable 56 degrees Celsius. Comparing this to the Intel stock cooler, it was a big improvement. The temperature the Scythe Ninja 4 produced also fell nicely between the other CPU coolers. The Intel stock cooler did not perform well under the tests; 80 degrees Celsius was not exactly a good figure for a processor to be operating at. All of the other coolers on the graph provided a good perspective on whether the Scythe Ninja 4 performed well or not. As you can see, it was quite competitive. It was not quite at the level of the Noctua NH-D15S, but the NH-D15S has a bigger fan, and it costs a lot more money. I am sure with another fan attached to the Scythe Ninja 4, it will deliver even lower temperatures.
For anyone who do not use headphones, the noise level of a PC is very important. I do not think anyone would like the sound of a jet taking off next to them while they are watching a movie, gaming, or working -- even if they were playing a flight simulator game. The subjective APH Networks scale we use to rate noise emissions is between 0 and 10; where 0 is complete silence, and 10 is a jet taking off. While idling, I would rate the Scythe Ninja 4 at 1.0. While under load, I will rate it at 2.5. This is confirmed with other staff members here at APH Networks. The fact there are controls on the fan itself to adjust the maximum speed can really help reduce the noise when the CPU is under load. Our tests were done on the high setting, so with Prime95 running, the RPM did really ramp up, but impressively, it was still quiet. While idling, the RPM dropped down to only 300, which made everything really nice and quiet. The Scythe Ninja 4 really lived up to its name in this regard.
Page Index
1. Introduction, Packaging, Specifications
2. Physical Look - Hardware; Installation
3. Test Results
4. Conclusion
