Best CPU Cooler – AM3 Review

Page : 1 2 3 4 5 6ALL


Test Method

As neither the Mega Shadow nor the Venomous X come with a fan, we strapped on a single 120mmx38mm San Ace 9G1212H1011 fan, widely considered one of the best fans on the market due to its under-volting ability (lower RPM) and high static pressure (perfect for heatsinks). This fan can produce up to 99CFM at full speed.

The problem we had with the NH-D14 was that Noctua supply 2 fans with the cooler, both 25mm thick. The issue was if we removed both and placed our San Ace on there, it would defeat the design the Noctua cooler was aiming for. We couldn’t use 2 San Ace fans as that would be rather unfair to the other coolers using 1 San Ace each. We thought about using the stock Noctua fans supplied, but they wouldn’t be able to compete with the superior San Ace fan. Thus we concluded that in order to make sure all our bases were covered, we tested the D14 using different fan setups.

Load testing gives the real indication of performance levels. For this we used OCCT and set each test to run for 30 minutes on each cooler. The maximum temperature was taken from CPUID Monitor. We also took into account RPM levels of the fans. We tested the fans on each cooler for idle and load at its lowest and highest RPM to see the difference in cooling performance (using a fan controller).

Idle testing was taken after an OCCT run had finished. We allowed a cooling off period of 15 minutes for the idle test. The minimum temperature was taken from CPUID Monitor.

We used both a stock and overclocked AMD system with a fresh copy of Windows 7 64Bit Ultimate. The Phenom II 705e chip was not unlocked as doing so would prevent us from monitoring temperatures within the Windows environment.

Cool & Quiet was disabled in the BIOS as well as any other energy saving features. OCZ Freeze was the thermal interface material of choice.

Moreover, we have also included a custom water-cooling loop for a quick comparison to aid in seeing how close the latest top air cooling solutions have come to low cost, decent water-cooling custom built. This kit consists of a Watercool HK Rev 3.0 LT widely considered one of the best CPU water blocks on the market. A single 120mm radiator, the RX120 from XSPC. The XSPC X2O 450, a pump and reservoir combo. This kit (plus tubing, fluid, 120mm fan) will set you back approx £130.

Test System

  • AMD Phenom II X3 705e @ stock 2.5Ghz w/ 1.2v & @ OC 4.0Ghz w/ 1.5v
  • Gigabyte GA-MA790FXT-UD5P
  • Sapphire HD 4670 512MB Ultimate
  • 2GB Super Talent Project X @ 1600Mhz 7-6-6-18
  • 320GB WD AAKS Caviar Blue
  • Corsair VX 450W w/ Sharkoon 1000 “Golf Ball”
  • Windows 7 Ultimate 64Bit

The results

All tests were run a minimum of 3 times (each run was followed by a remounting of the heatsink) and only best results are represented.

*Please note. Testing was done out of case. During our test session, the ambient room temperature was kept at 17-18°C.


If we take into account the Overclocked Load Test that gives us a real indication of how good a CPU cooler is, we can see clearly that the D14 is on top for the AMD platform. With all coolers using the same San Ace fan (even removing the 140mm centre fan from the D14), the D14 is ahead of the Venomous X by 1°C on high RPM. However, if you see our low RPM results, the Venomous X is 2°C better than the D14, but this will have something to do with the distance the San Ace has to push the air across the long heatsink without the aid of the centre fan.

At low RPM, it seems (even a 38mm thick, high static pressure fan) it can’t transfer cool air to the other tower, thus why Noctua insists on having the 140mm fan in place. We see this backed up when temps drop 6°C when you add the P14 pushing the air across to the other side. That said, when we returned the San Ace to high RPM and keeping the 140mm fan in place, temps were only 1.C better than without the P14 in place.

Our previous king, the Mega Shadow, couldn’t seem to keep up, which was a surprise. Even after 3 re-mounts/testing, it still wasn’t to the standard the other two coolers had set. Perhaps the Mega Shadow isn’t optimized for the AMD platform, just Intel.

When keeping the stock Noctua fans in place, we were very surprised to see the results it produced. Even if the D14 was using the two included fans, it was still 1°C better than the Mega Shadow using the high performance San Ace at full tilt, and 1°C higher than the Venomous X. When we reduced the fan speed using the low noise adapters on the two Noctua fans, temps went no higher than 38°C. Compare that to the Mega Shadow on low RPM hitting 41°C and the Venomous X 1°C higher at 39°C.

Be Sociable, Share!