|Introduction and Specifications|
|Test System and Temperature Testing|
Test System and Temperature Testing
Test Bench System
- AMD FX 8350 (Review)
- 16GB Kingston DDR3 Savage Ram (Review)
- GA 78LMT USB3 Rev 5.0
- Sapphire R9 285 OC 2GB
- Windows 8.1 PRO
Coolers included in the thermal tests:
When it comes to stress testing the CPU, nothing can torture the CPU more than Intel Burn Test (IBT). We decided to use this software to test the coolers performance. We ran 10 rounds of IBT at Maximum which is known to generate a tremendous amount of heat. The FX 8350 was put through three stages of temperature testing for each cooler tested. The FX 8350 was first run at stock speeds of 4Ghz, then overclocked to a reasonable 4.4Ghz and finishing off our testing and analysis with a final overclock of 4.7Ghz. The maximum temperature was recorded after 10 rounds for each frequency. HWMonitor was the software used for recording temperatures. Since it's common knowledge by now that idle package temps are inaccurate, we had no option but to record CPU temps for idle and package/core temps for load testing. IBT ran successfully through 10 rounds at all three frequencies indicating that the CPU has achieved a reliable amount of stability. We maintained an ambient temperature of 20c while testing these coolers. Our choice of Thermal Compound was the Noctua NT-H1 paste that was included with each of the CPU coolers. Noctua NT-H1 is a non-conductive thermal paste which is always good as you don't want to accidentally cause any damage to the area surrounding the CPU socket. We applied a pea sized amount and let the mounting pressure of the CPU cooler do the rest. The thermal paste was applied in the middle of the chip, where the most heat is generated. Below is a picture of when we removed the Noctua NH-C14S, clearly showing how the thermal was applied. We tried mounting the cooler again but the thermal was spread out much in the same manner.
Below is a picture of the base of the NH-D15S after removing the cooler from the CPU showing the thermal paste application.
The NF-A15 PWM fan included with the Noctua NH-D15S runs at 1500RPM. Noctua has also sent us additional fans for the purpose of this review. The NF-A14 PWM fan which came along with the Noctua NH-C14S also runs at 1500RPM. We decided to give the NH-D15S a go with a single fan in the middle of the Twin-Towers as Noctua's primary intent was reducing noise levels further by using a single fan with the NH-D15S. However, as the fans are so silent we figured that we would also test the cooling capacity of the dual heatsink cooler with two fans, both running at 1500RPM. The NH-C14S was also tested with two fans, spinning at 1500RPM. The NH-U14S and NH-U12S were tested with two fans in pull/push configuration with the fans at a maximum speed of 1500RPM. Although these are PWM fans, we thought it best for the fans to run at a constant speed of 1500RPM throughout our testing period.
Our first run is at 4Ghz, the stock clock speeds for the FX 8350. The idle temperature for all was at 23c. Idle temperature should always be taken with a grain of salt. The NH-D15S tops the chart with a 5c over ambient temperature. The difference between one and two fans for the NH-D15S is marginal and was quite expected. The NH-D15S with one fan performs similar to the NH-U14S with two fans running at 1500RPM. The NH-C14S performs exactly the same as the NH-U12S, topping out at a temperature of 28c, 8c over ambient temperature.
Heating things up at 4.4Ghz we notice how the margin between the NH-D15S running with two fans and one fan increases from 1c to 2c. The NH-D15S topping out at 30c is just 10c above ambient temperature. 4.4-4.5Ghz is the usual average overclock for an FX CPU. The NH-C14S performed very well, being just 5c hotter than the NH-D15S, making it 15c over ambient temperature.
Overclocking all the way to 4.7Ghz is where a difference in temperature between the NH-D15S and NH-C14S can be seen. The difference between them is 10c with the NH-D15S obviously being the better cooler, maintaining a maximum temperature of 38c. The NH-D15S is just 18c over ambient temperature at 4.7Ghz which is known to be quite a respectable overclock for any 8 core FX CPU. The NH-C14S was 1c behind the NH-U12S, reaching a maximum of 48c, which is 28c over ambient temperature. We do believe that as the heat generated from the CPU increases, the margin between all these coolers will increase even further. However, we were extremely satisfied with the cooling performance of the NH-D15S as well as the NH-C14S.