Articles: Cooling

Bookmark and Share

Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 ]

Performance Results in an Open Testbed

For our tests in an open testbed (outside the system case) we took Gigabyte GA-K8N Ultra-9 mainboard (BIOS F6). We used the same AMD Athlon 64 3000+ but working at higher Vcore setting of 1.7V (1.55V is the maximum ASUS A8N-SLI mainboard allows).

The lab room was 16 sq.m big. The room temperature during the entire test session remained at 21oC and was monitored with a thermometer. The mainboard was set horizontally.

We increased the CPU warm-up cycle from 15 to 70 minutes with the same S&M utility. According to the thermometer, the room temperature didn’t get any higher throughout the entire test session.

Here are the results:

Unfortunately, SpeedFan (version 4.23 and the latest version 4.25) would freeze on this mainboard at the sensors initialization stage that is why we used Gigabyte Easy Tune 5 to monitor the temperature.

Zalman CNPS9500 LED (click to enlarge)

Thermaltake Big Typhoon (click to enlarge)

The result obtained for Zalman CNPS9500 LED at 1900rpm is not accidental. The thing is that according to our subjective noise level estimate, Zalman CNPS9500 working at 1900rpm and Thermaltake Big Typhoon working at 1360rpm generate the same level of noise. At 2500rpm Zalman cooler is much louder.

Our open testbed tests allowed us to get to know Thermaltake Big Typhoon a bit better. Firstly, even under the maximum processor workload the heatpipes remain cold between the cooler sole and the first aluminum fins. However, the segments of the heatpipes right before the aluminum fins and right above them on top of the heatsink are hot, not just warm. The heatsink fins feel more like warm. Secondly, a have to correct our statement made above about the Thermaltake Big Typhoon also cooling the electronic components on the mainboard around the CPU. It does cool these components, but only in the idle mode . Under the workload such as during Super PI or S&M, the air flow coming down from the heatsink fins is warm and cannot serve any cooling purposes in the open testbed (not to mention the closed system case).

So, it looks like Zalman CNPS9500 LED may have one very last chance to regain its positions – Prescott. Will the efficiency of Zalman CNPS9500 LED in a Prescott based system make up for high price and high noise level?

Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 ]


Comments currently: 0

Add your Comment