Active Optical Cables - Rapid Market Growth Predicted
harry
April 7th, 2009
Comments (0)Two news reports, within a week or two of each other, cause me both excitement and concern.
The first, a report by Information Gatekeepers Inc. highlights the recent very rapid increase in sales of “Active Optical Cables” and makes predictions of dramatic future growth.
The second reports that EMCORE Corp., has received a patent award for its Active Optical Cable technology. US Patent No. 7,494,287 B2 with “…broad claims covers all fibre-optic active cable applications…”.
Of course the industry is not short of reports that claim exponential future growth for almost every conceivable technology. In the past most of these seem to have been based on minimal actual data and maximal imagination. However, this time the claims seem credible.
The concept of an “Active Optical Cable” is so simple it is almost trivial. You integrate optical transceivers into the ends of an optical cable and provide standard electronic interfaces to connect to user equipment. With recent improvements in optical transceivers this has become relatively easy. At very high speeds (above 10 Gbps) Copper cables are problematic even in short distance applications. Fibre is lighter, easier, offers very greatly extended distances and solves the problems associated with “ground loops”. Most existing equipment has copper interfaces built into it so an AOC can just “plug-in” without change to existing equipment.
Initially applications are for relatively short-distances (a few 10’s of metres) such as interconnection of high-performance computers, servers, routers etc. within a central computer complex. Such high-performance complexes are no longer the sole preserve of the pure R&D and scientific world but are increasingly being adopted by businesses for applications such as simulation, modelling and computer-aided design. However, one supplier is reported to be targeting longer-distance applications and is reported to have announced AOC’s of up to 3km in length. Of course, you don’t really want to “pull” a cable that has bulky integrated equipment on its ends through a long and convoluted cable-way either. Importantly, AOC’s can mitigate the reluctance of many users to install fibre. To many users fibre seems difficult to use and if familiar copper will do the job then why change?
Someone designing an AOC can design the whole link (transceivers, fibre, connectors etc) to optimise its cost/performance without needing to be constrained by “industry standards” or what other suppliers are doing - all you have to do is conform to the electronic interfaces at the cable ends. This should encourage rapid innovation and development. A potential disadvantage is that centralised Network Management systems do not get access to detailed performance and management information (error rates etc.) as seen on the optical segment of the link. At the distances involved this may not matter.
The worrying part is the “broad claim” of a patent. If this patent is for the detail of a particular implementation then it should be applauded. If it is for the overall concept then we have yet another example of the patenting of a very obvious generic principle. The principle of replacing an “electronic” link with something else even by use of devices integrated within the cable has been in wide use for many years. I have not been able to find sufficient detail to decide. But it’s a worry.
Of course the AOC market may well have a limited lifetime. When computer-room equipment becomes routinely available with optical interfaces built-in then advantage of AOC’s will vanish. In the meantime which may well be 5 or more years there seems to be a market there.


Ralf Haller 








