440BX Motherboard Review - Fall 1998

Supermicro P6SBA

• BIOS Version: 1.2B (July 15, 1998)
• Board Revision: 1.01
• External clock speeds: 66, 100 MHz
• Slots: 4x full size PCI, 3x full size ISA, full size AGP
• Memory: 3x DIMM

This board uses the AMI Winbios which is really something you need to accustom to. I have to admit that I don't really like it; but it allows the resources to be manually distributed. The jumpers on the boards are all in colours of the American flag, but they are still jumpers! Four for the multiplier and one for the front side bus: 66 or 100 MHz can be selected, no more. The BIOS has one heavy deficit: On the one hand it is possible to disable the USB controller, on the other hand the IRQ is still casted by it. That explains troubles you could face while trying to install multiple IRQ-requiring expansion cards.

The memory tests were all successfull except the 32 chips module which wouldn't run at all. But I cannot say this enough: This DIMM doesn't have to run at 100 MHz, but it is always interesting to see if a board runs with such "bad" memory anyway. The board allows three fans to be connected, one for the CPU and two others for the chassis. I expected this board to be one of the faster one's since it achieved quite good benchmark results in Tom's last BX review in spring. But it seems as if Supermicro couldn't improve the performance at all since it only scores average results this time.

All in all there are too few features and too little advantages to recommend this motherboard. An option to release the USB IRQ is obligatory; I would gladly abandon the Winbios for it...

Superpower SP-P2BXA

• BIOS Version: B.3 (August 4, 1998)
• Board Revision: Ver. B
• External clock speeds: 66, 75, 83, 100, 103, 112, 133 MHz
• Slots: 4x full size PCI, 3x full size ISA, full size AGP
• Memory: 3x DIMM

I did not know Superpower before this test. While I was unpacking the board for testing I noticed that one IDE connector was yellow instead of the dominant white colour. I first thought this was a cosmetical error, but it is on purpose: The second IDE connector is yellow to be able to find out quickly which one is which. In a dark and thrived case it's usually almost impossible to read the little prints on the board.

The CPU setting requires manual work. Some dip switches want to be set for the wished multiplier and one double jumper determines wheter the system runs at 66 or 100 MHz FSB. There are overclocking settings available for both speeds; they can all be selected in the BIOS. Unfortunately there is missing a BIOS item to disable the seizure of an IRQ for the USB, so you will have one less even if you don't use the USB (I think the right name for USB should be "unused serial bus". It's nice to see that a newcomer board reaches a middle place in the benchmarks. All memory I threw against it ran fine, except the big 64 MB type (PC-66).