Testing

To test the Sylphee, I’ll be using an AMD Athlon 64 X2 3800+ with Arctic Silver 5 between (300 hour curing time elapsed) sat in a Lian Li PCB-20 case. To test, I’ll use the following process:

Let the core idle for 30 minutes with nothing running, and record the end temperature.
Start both StressPrime on both cores and a single instance of folding@home.
Let the testing run for 30 minutes and record the end temperature.

Ambient was a constant 19C.

While in use, the Sylphee stayed well under the audible radar, and with the case closed this cooler can be classed as silent. Overall its performance is good for its noise level, although I would be reluctant to overclock with this cooler attached. Managing to cool near to the mark the ThermalTake Max Orb set at idle is no mean feat. It manages to beat the stock cooler hands down and produces a lot less noise in the process.

During testing, the Sylphee never reached its full speed of 3600RPM, but hit 2200RPM max. This goes to show that my earlier point regarding thermistor positioning is valid. Placing these temperature sensors nearer the core would provide more useful temperatures. It also shows that there is still some cooling headroom left in the Sylphee, allowing for more powerful CPU’s to be used at the cost of noise.

As the fans blow outwards from the heatsink, they also provide a small amount of airflow for the surrounding components. While no official testing was carried out, the four sticks of OCZ 512mb DDR2 RAM felt considerably cooler while using the Sylphee.