FAC, nice digest, here's another tidbit on Lynx
theneteconomy.com
Rethinking Optical Switching
All those ports are just marketing fodder so far
By Joe McGarvey
When it comes to building the first-generation of optical switches, which will divert wavelengths of capacity from one end of the Internet to another, startup Lynx Photonics is thinking small.
In opposition to established equipment makers and a handful of startups that are building optical switches that can handle thousands of wavelengths of capacity, Lynx is focused on building the core switching fabric for devices that will handle between eight and 64 wavelengths.
"The sweet spot is in this range," says Michael Leigh, president of Lynx Photonics, which this week unveiled portions of its product roadmap. Leigh also said that several equipment makers were currently evaluating the company's first product, a photonic switch based on so-called planar waveguide technology. This first module is capable of handling 8 wavelengths of optical capacity.
Despite the publicity surrounding the construction of optical switches that start at about 1,000 ports from Calient Networks, Nortel Networks and others, Leigh says that carriers will be reluctant to deploy switches of such huge capacity in their networks, at least not until the technology is proven. These large all-optical switches, which are expected to be available sometime next year, would be placed at large switching centers and would be responsible for multiple terabits of traffic reaching its destination. Leigh argues that few carriers would be willing to trust such a high percentage of traffic to a single device.
Instead, says Leigh, carriers will take some of the risk out of deploying optical switches by distributing the switching functions to smaller switches scattered about their networks.
"Carriers are looking for a distributed optical switching approach," adds Leigh.
The major difference between Lynx's technology and switches that can scale to 1,000 ports or more is the underlying technology. The Lynx switches are based on waveguide technology. By carving a tiny grid system into a silica block embedded on a silicon semiconductor, the Lynx switch can alter the path of wavelengths by diverting the laser signal at any of the intersections in the grid system.
The vast majority of large-scale switches that have been proposed are based on tiny tilting mirrors, which can be manipulated by tiny mechanical devices built into the silicon to reflect light to an outgoing port. While acknowledging that three-dimensional mirrors offer greater scalability than waveguide approaches to building switches, Leigh says that the technology embraced by Lynx is closer to implementation that systems based on tiny mirrors.
In addition, he says that the waveguide manufacturing process is much more sophisticated than the mass manufacturing of systems based on tiny mirrors. Perhaps the most overlooked aspect of the construction of all-optical switches is the ability to assemble finished products on a large scale. Even the most promising technology is useless, says Leigh, if it cannot be advanced past the prototype stage.
Lynx is focusing its first products at system makers that will build small-scale optical cross connects and optical add/drop modules for the metro environment.
=====
Do you think that pursuing a *think small* strategy would eventually catch up with them? I would think carriers,once the toe is in the water, would want to scale up out of lynx's niche. They've got to keep that cost of ownership down.
regards,
dkg |
