The last special feature of the SB81P barebone is the chassis. The SB81P is the first mini-computer from Shuttle to use the fifth-generation (type P) chassis. Besides being assembly-friendly, this chassis is designed to optimize the airflows inside the system for better cooling and quiet operation.

Air comes in through the vent holes in the sides and bottom of the barebone. The space inside the system case is divided into three subsections: the CPU section, the section with the hard drives which are outside the basket, and the rest of the internal space. Each section has its own stream of air – that’s why the air doesn’t have time to get hot. The CPU area is cooled by the ICE module I’ve already talked about. The air stream created by this module goes across the case and never mixes with the other streams. Air from the other vent holes is divided into two streams which cool the remaining areas. The hot air is exhausted from the case through the power supply’s fan. The HDD section is cooled by two 60mm fans located at the rear panel, above the power supply.

They are connected in parallel to one fan connector and are recognized by the system as one fan.

That’s how the technologies mentioned above work. My task now is to determine their efficiency. I didn’t find a Windows-based monitoring utility among the software supplied with the SB81P, so I had to use a standard toolset consisting of S’n’M for loading the processor up, 3DMark05 for loading the whole system, and SpeedFan for monitoring the system’s status. I performed my tests at 200MHz FSB (regular mode) and 214MHz FSB (overclocked mode). The testbed was configured like follows:

• Pentium 4 560 processor (Prescott core)
• XPC SB81P barebone
• Integrated graphics core (the system load is the highest with it)
• 2x512MB OCZ PC4200

The results of the tests are listed in the following table:

And this table shows you the speeds of the fans in different tests:

Not bad for a barebone. The CPU temperature only grew up by 11°C under load – that’s good. And the result almost didn’t change in the overclocked mode – that’s excellent. The temperature in the system case remained almost the same throughout the tests, proving the highest efficiency of Silent X technology.

As for the noise the working SB81P produces, there’s not much to talk about. When the system isn’t fully loaded, the barebone is almost silent. Its barely audible hum isn’t annoying at all. When there’s a 100% load on the computer, the noise is somewhat louder, but remains acceptable nonetheless.