Intel confirmed key aspects of the long-awaited USB 3.0 specification last week, but the hugely improved performance is perhaps the least interesting of them.
Of course, USB 3.0 - also be known as ‘SuperSpeed USB' - will have a theoretical throughput of 4.8Gbits/s, ten times that of today's USB 2.0, a useful step up in performance. That assumes that by the time it appears in the 2010 timeframe, storage drivers have been rewritten to take advantage of it, but that's probably a given.
The software writers will also have their work cut out dealing with the technology's other differences from USB 2.0, with which it will, mercifully, be backwards compatible. USB 3.0 is a two-way design, with a re-engineered communications protocol designed to manage multiple data streams at once, a huge change compared to the naïve broadcast approach of USB 2.0.
The cables, which will be somewhat thicker than today's, will also have the ability to carry more power, allowing USB to expand its potential as a way of charging a range devices. The day of the charger could be nigh. Using a signalling protocol based on polling for active devices rather than broadcast is also claimed to save on the power drain exerted by USB on its host.
This is just as well because the number of devices that can be attached through a single controller will expand from today's limits. If there's ever a version 2.0 of the 3.0 incarnation, it seems likely that the interconnect will become optical.
The known unknows? How far will the average USB cable reach? What sort of price premium will the technology attract? Will any of this effect the rather halting progress of Wireless USB?
As great as ‘ten times the performance' sounds, it won't look as good by the time it arrives. Video is already heading for HD territory, not to mention the ever-expending size of image files as camera megapixel counts climb. Performance is also always relative to the application it is being thrown at - just ask anyone who uses Vista to do anything.