By now you're probably thinking 10GBase-T could save you money while speeding up your overloaded backbone, and in theory it should. You might already be using fibre-optic versions of 1Gb Ethernet successfully on the backbone or around the campus, and upgrading to 10Gb should only be a matter of changing a few cards in the switch or router, after parting with the necessary dollars. However, it might still be some months before you can realise any savings by attempting the same thing with copper, despite the inherently lower price of the cable plant.
For starters, you will actually have to get new cable pulled if you want to achieve the maximum 100m distance, but before that's done you'd better check the specs on your existing chassis to make sure it can handle the new speeds. A lot of early chassis backplanes were only designed to aggregate 100Base-T feeds into a 1Gb uplink, while later models which were built to handle 1000Base-T feeds were only expecting to drive 2-3Gb fibre uplinks. To successfully handle 10Gb uplinks the backplane could need in excess of 1Tb of bandwidth, which many vendors are now offering.
Your next problem could be waiting for the manufacturer of your particular chassis to make 10GBase-T ports available, and affordable. The market leaders for shipped 10Gb Ethernet ports at August, 2006 were Cisco, Force10, Foundry, Nortel, Extreme and HP's ProCurve, according to analysts at Dell'Oro Group, and although those ports were predominantly fibre optic-based, we can expect these vendors to lead the way with UTP solutions. The cabling companies all claim they are ready, but waiting on the electronics.
If, as is widely predicted, the history of previous Ethernet speed hikes repeats itself, we can also expect entrepreneurial start-ups to appear, chasing the high-speed market with innovative and lower-priced solutions based on the next generation ASICs from suppliers like Solarflare and KeyEye. However, fully tested, shipping, interoperable products could still be as far away as mid-2007.