I'll happily concede, but can we agree that the more honest specification would be the "max interface transfer rate" of 100MB/sec in Ultra DMA mode-5?

Well, that's 100Mbytes/s, don't forget, or 800Mbits/s. You would have to have each drive on a separate ATA100 channel to saturate the 500Mbits/s head rate on both drives at once, but the Mark 1 had two IDE channels, and while we're dreaming let's dream in high bandwidth. You can keep the five-spot, or give it to the EFF or the World Wildlife Fund or something.

500Mbits/s is a hell of a lot off a 2.5in drive, of course; I've got a 15K U160 SCSI drive here that's only 640Mbits/s head rate. And that theoretical 80Mbytes/s ends up as about 45Mbytes/s through the filesystem, so perhaps we can't quite saturate GigE onto our two drives. We can still go an awful lot faster than 100Mbit allows for, though...

Peter