Siemon 10 gigabit capable network connector

A recent workshop held between some of the major network cabling, connector and chip manufacturers to discuss the development of  copper networks beyond 10Gb/s was held at Pennsylvania State University CICTR (Center for Information and Communication Technology Research) in August last year.

Why I have only just recently received a press release from one of its co-sponsors, Nexans, is anyone’s guess, however I thought it relevant that I should share the outcomes of this workshop held between some the finest technical minds, and industry leaders in the manufacture and development of copper networks; we are always keen to know what the next big leap in speeds will be. Currently the maximum speeds attainable on both copper and fibre network systems is 10Gb/s, which some would say is fast enough; however we believe that if high definition TV, HD audio and 3D TV continues its development and mass-adoption at the pace we are currently seeing, it won’t be long before even 10Gb/s will seem like a very slow network indeed, even in domestic settings. If we consider that blu-ray movies can be 50 Gigabytes in size, and soon 3D titles could very well require more that 33% more capacity, it is not hard to understand that the transfer, backup and streaming of high-definition video content is going to put demands on networks never seen before on the internet or within domestic settings.

In essence the workshop concluded that in 2010 there is sufficient cause and interest in creating a committee to call-for-interest to establish a 40BASE-T solution.

The key observations, which I have copied verbatim from the press release are:

  • Copper solutions beyond 10G could include many possible options but the highest Twisted Pair Category (Category 7A) is clearly in prime position.
  • Data Centre architecture will play a vital role in next generation networks. Jeff Cain, of Cisco systems, suggested that rack-centric, modular designs could change current thinking.
  • Chip Vendors and Data Centre designers pointed out that structured cabling had many benefits over ‘Top of Rack’ switching, and that future copper solutions for 40G shall not be “point to point”.
  • The concept of a unified I/O, spawning the prospect of “any machine, anywhere, anytime” was introduced.
  • The PHY vs. Channel, aka “Chicken or Egg” debate, continues…with chip vendors saying the cable channel needs to be established first, while the cable manufacturers say the PHYs need to be developed first. Workshop attendees agreed that concurrent development was the best plan.
  • Based on his market analysis and model, independent consultant Alan Flatman (LAN Technologies) suggested that the call for interest (CFI) for 40GBASE-T should be initiated in IEEE during 2010.

Bottom line is it is actually possible for you to install 40GB/s capable networks today, even though this technology has yet to be developed. Category 7A cabling systems are already available, and if these cabling systems are installed correctly they will be able to deliver speeds well in excess of 10gb/s, where some cabling and connector manufacturers are even guranteeing 40gb/s performance, even though such technology is only in conceptual stages.

TEA London are already installing Category-7A networks, our first cabling project of this nature started in early 2009, following acceptance of a cabling design proposed in 2008; such cabling systems are not significantly more expensive than good Category-6A copper networks, and if one takes into account that a Category 7A network is truly future proof, and will likely give a lifespan of a minimum of 15 years, but likely as much as 25 years, then this is an investment well worth making in any new build, or during any refurbishment project where you own the building.

The more common Category-5e cabling systems are less expensive to implement, but is already at its end of life, where you will be hard pressed to get a stable optimised 1GB/s network, let alone 10Gb/s, similarly, cheaper Category-6 networks have proven themselves to only just scrape through in passing minimum requirements.

We have been advising our clients to implement a minimum standard of Category 6A cabling systems for some time now, you should strongly consider the same if you consider what developments are around the corner.

Should 40Gb/s copper networks appear in the near future, here is an important fact worth considering: it will be faster than fibre optic networks, currently copper 10Gb/s networks match fibre optics (and we mean only high-0end fibre optic networks, over short distances, using the most expensive equipment), the key difference being that copper networks are always cheaper, in terms of hard costs, implementation costs and installation costs; fibre optic networks have the advantage of being able to deliver data over long distances reliably, but copper networks will soon take over as the faster technology.