The post a few weeks ago about tests comparing shielded and unshielded twisted-pair cabling systems' ability to handle 10-Gigabit Ethernet generated a bit of feedback in the way of emails and phone calls. One commenter pointed out that an article we published two-and-a-half years ago predicted that such tests would produce the results they did.
More recently, a quick trip around the internets revealed not one but two videos showing how quickly a shielded cable can be terminated - in each case less than a minute. Take a look at Tyco's video demonstration and/or Siemon's.
Meanwhile, my editorial in the December issue of Cabling Installation and Maintenance touched on the notion of running more than 10-Gbits/second on shielded twisted-pair cabling. First, some make-good housekeeping. In that editorial I neglected to mention that the in-the-works 40/100-Gbit Ethernet specifications within IEEE do indeed include a twinax copper-based solution. In one of the many instances in which my brain took a vacation while my body was working, I stated that the 40/100-G specs are fiber-only. D'Oh!
Moving ahead, if I possibly can, I promised more detail on the possibility of 40/100-G over shielded twisted-pair. This article from the January issue of our magazine provides some of that detail. As the article indicates, it looks like it could be a long road for proponents of an IEEE-produced 40/100-Gig-over-shielded-twisted-pair specification. From the looks of it, ensuring that a shielded cabling system can handle such throughput rates might be the lowest hurdle to overcome.
We'll continue to track what goes on with the possibility, and keep you informed.
Subscribe to:
Post Comments (Atom)
1 comment:
When terminating, surely it it more important to make the correct connections rather than just do it quickly?
In the Siemon video the pairs are inserted 1&2, 3&4, 5&6, 7&8 so no wonder it was done quickly!
I'm not aware of any standard that uses the pinout shown in the video.
Whilst Cat7 doesn't have to conform to EIA standards, I was under the impression that RJ45 8P8C used E568A/B layout.
Even RJ48 & RJ61 does not put the pairs together.
So, is the test valid?
I suggest that they try it again in accordance with TIA/EIA-568-B.
Post a Comment