FIC also offers an ATX board with the Apollo Pro chipset, but the sample we got is the baby AT version. It is a board crowded with connectors, chips, jumpers or other parts to make the board's size as small as possible. In this case I think the components are so close together that it's almost impossible to plug in the PS/2 mouse port cable if the CPU is already installed. The clear jumper instructions which FIC usually prints on the board do suffer from this.
Even a soundchip (Yamaha) has been placed on the board. The idea FIC had was certainly good, but using a board which is one or two inches longer would improve this lack of space. Due to this, the board comes with three DIMM sockets, but also only three PCI slots. Here I have to criticize two things: First it is very difficult to plug in a long PCI card, because the IDE connectors are both right in front of the PCI slots. All people who use both ports will have to fiddle around with the cables to arrange them properly. Second the board's IRQ assignment system is not very advanced; every slot has its fixed IRQ. The PCI slot one does use the same IRQ as the AGP slot, and in slot three my Adaptec controller collided with the IDE controller. In slot two if worked fine, but than there's no chance to run two Voodoo² cards in SLI mode on this board; the only way would be getting a very large SLI cable (I don't know if that's a good idea) or using a Voodo² card with two chipsets on one board (e.g. Quantum 3D X-24).
The memory tests were passed mixed: The LGS 7 ns PC-100 DIMMs ran fine, but the Toshiba PC-100 memory ran fine only if just one module was installed. Two modules lead to exception errors and blue screens. The 256 MB module did not run at all, but then again the 32 chip module with LGS chips (PC-66, 10 ns) ran properly as well as the PC-66 memory from Samsung.