Well, despite my best intentions, time in customer and industry meetings once again outnumbered time spent in technical presentations at ECOC 2012 in lovely Amsterdam.
I did manage to catch some of the presentations on silicon photonics and it’s clear that this area is continuing to get a better understanding of both the underlying technology’s capability and it’s limitations. This technology leverages the incredible manufacturing base which has been developed for the IC industry and so researchers have access to state-of-the-art fabs capable of producing commercial-grade devices even for research projects.
However, the relatively slow commercial uptake of silicon photonics to date might be indicative of the fact that, for all it’s promise, Si photonics isn’t yet able to deliver the step improvement in size/power/performance that its advocates suggest.
The Post Deadline papers are now available (http://www.ecoc2012.org/documents/ecoc2012_postdeadline.pdf) with the usual array of spectacular results. After a quick look through, the standout paper for me is by Takara and co-workers from NTT where they achieve a spectral efficiency of over 90bits/s/Hz in a multi-core fibre for a total capacity of over 1Pbit/sec on a single fibre. Very impressive numbers from a ‘hero’ experiment but it’s still unclear to me that multi-core (or, for that matter, multimode) technology will ever be deployed in a full commercial system.
The beer at the Finisar customer event at the Heineken Experience last night was good, as was the company (thank you Stuart, Ian and Steve) and I’m looking forward to one more day here before the loooong flight back to Oz.