Biostar absolutely and utterly fails at BIOS and insists on ****ing around WAY too much with the reference builds. End result? They break ****. Like this. And they go "oh well everyone uses X so why bother fixing it?" and ship it anyways. If it wasn't supplying enough power, it wouldn't work period. The end. Biostar's crapfest simply doesn't init the LPC SIO correctly until the protected mode transition - this is an unequivocally broken BIOS. This is NOT the first time I've seen it either.
TL;DR: If you want a working system, you need a new motherboard. From a different manufacturer.
That makes sense, I was wondering why it worked so well once it finally gets past the initial startup. It was a really cheap motherboard and I've already found that the onboard NIC is pretty terrible too, I guess I'll try to upgrade it to something better when I get a chance. In the meantime I'm going to get one of those cables, thanks for the recommendations everyone. Since PS/2 is going to be phased out eventually I'll need it someday anyway because I can't imagine typing on anything but buckling springs now. 
Yup. And Biostar ain't gonna fix it.
Ever. So yeah. Chuck their garbage. They're one of the few vendors to make my blacklist for BIOS stupidity.
The most hilarious part is that 8 out of 10 of their BIOS defects that would automatically FAIL any sort of certification process? ARE SELF-INFLICTED. Baseline is JUST fine, and then their idiots get their hands on it, break it horribly, and ship anyway. And you can outright forget ever seeing a fix. Hell, they don't even do their microcode updates on time, much less correctly. Upstream basically GIVES them exactly what they need to do, and they manage to screw it up anyway.
Ten bucks says the morons pushed the SIO's IPL to protected mode ROM segment
again because they put the bloated and permanently buggy Atheros POS (seriously, I wouldn't use the 8152 to torture people) in real-mode for PXE and WoL that doesn't even #$@&*ing work.
