I rather doubt USB has any more slack for hacking it up again. “thought the whole idea of USB was to reduce the amount of ports, how about putting effort into making USB faster, rather that special ports for storage devices? ” USB won’t die any time soon, but handling external hard drives is not something it’s terribly good at, compared to the bandwidth needs of mice, wireless NICs, etc. Some folks are already using external SATA drives, as well, and having a more rugged physical interface would not hurt if you’ve already been sold on the use of SATA. Given the cost and convenience of external drives (especially if you want data or burning features accessible to multiple PCs), this may be a nice step.
It also offers a nice alternative to non-USB techs like Firewire and ATAoE for short-distance use (remember, FW has royalties involved–it’s kept Firewire from massive adoption similar to USB).
Some USB chip combos (the host and bridge matter) can break 35MB/s…and use a fair bit of CPU power to do it.ĮSATA should have CPU saving and compatibility benefits (or at least lack of occasional data corruption for no apparent reason) over common USB chipsets with the caveat that you need the port (so not compatibility in terms of randomly choosing a PC< but randomly choosing a drive to use). …but they can break ~30MB/s of USB/Firewire, some now going over 70MB/s (the data isn’t too big).