In order to compare the relative performance of the Battle-Axe, I tested the idle and load temperatures of both the core and PCB using the NVIDIA reference cooler as well as the popular Thermalright HR-03 GT.
All temperatures were observed using RivaTuner V2.11. To test full load temperatures, I used the Folding @ Home V6 GPU2 client. Why not help cure cancer while reviewing a VGA Cooler? To obtain each load temperature, the GPU2 client was left to run until the temperature peaked and remained consistent for a minimum of 20 minutes. The ambient temperature was maintained at a constant 22°C.
A 92mm Panaflo fan was used on the Thermalright HR-03 GT with an airflow of 42.7 CFM.
In order to adjust the fan speed of the dual fans of the Battle-Axe and Panaflo of the HR-03 GT, a Zerotherm fan controller with a manual dial was used. Low corresponds to the lowest setting the dial would allow, medium the exact middle of the dial, and high the maximum the dial would allow.
Immediately apparent is the drastically lower idle and load temperatures afforded by using an aftermarket cooler in favor of the NVIDIA reference. The Battle-Axe is narrowly edged out of first place by the HR-03, however this small margin could very well be due to the two less ramsinks applied during Battle-Axe testing.
Core temperatures are where the Battle-Axe really shines, quite possibly an indication of the effectiveness of Xigmatek's H.D.T. Again there is a small margin between idle core temperatures. Load core temperatures are another story. The Battle-Axe manages to keep the core 7°C cooler on high, and 3°C cooler on medium!
Noise
While I have no actual decible meter, it was more than simple to compare fan noise between the Battle-Axe's dual 92mm fans, single 92mm fan of the HR-03, and NVIDIA reference.
As the majority of us can attest to, the NVIDIA reference fan is only inaudible at the auto setting, while being commonly referred to as a "fan blower."
While my case is not the most quiet around with 3x Low Speed Panaflo 120mmx38mm radiator fans, a Swiftech MCP-655 pump, and two 120mmx25mm case fans, even when set to High the Battle-Axe added no noticeable increase in noise to the system.