This Is How We Tested

For this test, we looked at the loudness and heat generation and not the speed measurements of the motherboards. We used an Intel Pentium 4 3.2 GHz for the Socket 478 or alternatively a P4 3.6 GHz for the Socket 775. For the AMD versions, we used an Athlon 64 3400+.

The assembly of the hardware and barebone was also evaluated as well as the cable routing and the edges on the mini PCs.

To measure the sound level, we used the SL-4001 from Lutron, held at 50 cm from the barebone. The distance was set to 50 cm because the barebone is usually put on the table and is consequently not more than 50 cm away from the ears. Two measurements were taken. One was taken when the barebone was in idle mode and one was taken after approximately 30 minutes with the CPU under a full load.

We used the FAN-O-Matic from Innovatec for the temperature measurements. Many heat sensors were attached to the barebone for the temperature measurements. A sensor was placed in the cooling fins of the CPU fan, one on the hard disk, one on the DVD drive and one outside on the power supply. For each barebone, the measurements took three hours, during which the onboard graphics was used. With the test program Prime95, the CPU and the motherboard were heated up or alternatively loaded.

The scope of delivery and the configuration of the barebone were also included in the evaluation.

Test Equipment

We equipped the barebones with the hardware shown below. This included an AGP and a PCI Express graphics card to replace the onboard graphics. To ensure that the graphics card fans didn't affect the decibel measurement, the noise measurements were only taken with the onboard graphics. If the barebone didn't have onboard graphics, like the Soltek QBIC for example, a passively cooled graphics card was used.

Small fan: makes noise like a large one

The operating system is on a Seagate S-ATA disk

The Corsair dynamic duo: Should be 2 x 512 megaBytes

5.25" drives have different depths. Things that are unimportant in a midi tower can cause installation problems in a barebone. Therefore, another item on our checklist was how well the hardware installed and where the weak points were. In the process, two additional items were checked: how is the cabling and the workmanship of the barebone case?

Every centimeter can save space for the cabling