Performance

Performance of a cooler is affected by several qualities of the cooler. These qualities are heat transfer (which is a result of flatness, pressure, and material), and how fast the unit dissipates the heat it has gained. With water cooling, a bigger radiator, with larger tubing and a more powerful pump will help dissipate the heat faster. A flat high quality water block will ensure that the maximum amount of heat is conducted.

Before going to the benchmarks, let's see how flat the contact surface of the block is. For this we used the old knife test trick, and simply seeing how well the contact was. The block was put on, then taken straight off and we inspected the contact.


There was some light shinning through when we did the knife test... However, when we went to remove the block, it was quite hard to get off. After finally getting the block free, it was easy to see the contact was excellent.

To see how the Silent Water II performed number wise, tests were done using an AMD A64 4000+ E4 revision (San Diego core). Test were run at stock speeds, then with increased voltage, and finally with a overclock. The fan on the Silent Water II was also set to high and low during the tests. When it came to overclocking, the highest possible clock that I could boot Windows with the stock cooler was 2.76 GHz. It did pass SuperPi runs, however it failed Prime95.







The numbers are quite impressive for this water system considering the specifications of it. The CPU was clocked to 2.8GHz 1.61v as well, which was Prime stable. I also managed to get the system into Windows at 2.9GHz, but this was not Prime stable. This particular A64 is a voltage hog and I didn't feel comfortable going much higher than 1.65v with this cooler...