Environmental photonics: Is the glass always greener?
OFC / NFOEC 2012 will host a great programme on Green Photonics. With sessions designed to stimulate discussion amongst delegates, permit me to contribute my own thoughts to the debate.
First, environmental considerations have already overcome the hyperbole barrier, at least here in Europe. Incumbent telcos have implemented carbon reduction programmes and many are amongst the largest (proportional) users of renewable energy in the region. Green awareness is particularly acute within the cloud computing domain where many (though sadly not all) software-as-a-service providers have secured low-carbon energy sources and righteously advertise such credentials. In theory, low-energy proposals should be well received by network operators though I suspect commercial considerations such as supplier base security remain the overriding factor.
Second, whilst supporting judicious use of government finances to bridge funding gaps that have starved photonic innovators over the past few years, I believe we should avoid over-hyping the investment potential. High-temperature semiconductor optical amplifiers and low-energy silicon CMOS photonics justifiably, in my opinion, have received federal funding. Yet with investor appetite for optical networking starting to thaw, we must be alert to the dangers of endorsing unrealistic growth expectations. Between 2003 and 2010 over $3 billion of shareholder value was destroyed in the pursuit of portable fuel cells. The same must not happen with Green Photonics.
Third, I believe it is important to quantify the end-to-end environmental merits of optical networking proposals. Failure to do so may instil the photonic equivalent of off-balance sheet accounting with equally misleading consequences. For example, would a 40-Gb/s MLSE chipset consume more or less per-bit energy than 4×10-Gb/s transceivers with in-line dispersion compensators? Automotive engineers already understand these considerations and have codified environmental impact formulae such as well-to-wheel efficiency. Similar terminologies exist in the photonics domain, for example wall plug efficiency, but these should be expanded to encompass the entire photonic journey.
Fourth, I believe circuit switching will have an important role to play in reducing network energy consumption. This is especially true in the transport layer where entire routes could be hibernated according to predicted loads fluctuations. I mention this example because it illustrates the growing need for joined-up thinking between component designers and network planners.
Dr David Nugent is Founder and CEO of Elucidare Limited, a boutique technology development and investment advisory business.
Posted: 27 January 2012 by
| with 0 comments
Filed under: cloud, computing, consumption, energy, Green, network, Photonics
Blog post currently doesn't have any comments.