Today I was given a "sunPCi" possibly IIpro. Not really sure what model it is, as I left it in the lab at the university. (A professor was cleaning out his office because they're moving him to a temp location while they remodel things and randomly gave it to me as a joke, woot for new lab!) The thing is mostly useless without a Sun workstation, and the university got rid of the last of those years ago, but I figured there might be some chance someone here might have used one and had some ideas. These cards, a massive PCI card, is a 'complete' computer on a board. The idea was to enable SUN system users to run dos/windows systems in native i86 processors in the same box as their SPARC processor based systems. I'm thinking it might be possible to mount this thing in a custom built PCI bracket with a small power supply and rig up some way to run a file server. I was told it had a 700mhz celeron or something rather weak like that. I've been reading random stuff on these cards, and most seem to lack any hard drive connectors, instead relying on the main drives of the mother system or a network drive for storage. The one I had did appear to have an IDE style connector, but I'm not sure if that is IDE or whatever it was that floppy drives used. (It has been far, far too long since I had to deal with a floppy drive from that era to remember if it was its own connector or IDE.) Any thoughts?

It needs the host Sun machine to feed it with data and code. It runs special drivers that allows it to do I/O through the host. You basically have a paperweight on your hands.

Sadly it isn't even heavy enough for its size to work as a decent paperweight. Still might be fun to play with even while expecting to fail.

Keep it around for a couple of years and then flog it on eBay to universities when Sun is long since bankrupt.

More seriously, if you can root out a copy of whatever Sun uses to load the thing with data, you could probably disassemble it (or look up the source code if it's open source, which I doubt). Once there, you might be able to rig a PC-side driver to send that necessary initialization data and communicate with it further. This is like a full-time project, however, so you should take advantage of your professor being distracted by the move to pursue it.

Power isn't nearly as hard. I believe there's only a few pins on the PCI slot that actually handle power, so just wire those babies up and be careful not to blow it.