It's never too early?
by Stephen M. Hardy
Last month's NXTcomm08 in Las Vegas reminded me of one of the optical communications industry's salient attributes: It's never too early to tout a technology or product.
I'm not talking here about our space's well-earned reputation for announcing products well before they're ready. True, this trait isnâ��t as pronounced as it was during the bubble. But the fact that I still have companies in this space showing me technology demonstrations "to prove it's not vaporware" indicates just how common the sin of premature announcement continues to be — and how aware of the practice members of the community remain.
Last month in Las Vegas there undoubtedly were instances where companies listened more closely to their marketing departments than their engineering cadres when it came to the timing of product announcements. But what struck me more forcefully at NXTcomm was how quickly companies leaped to demonstrate standards-driven technologies well before the standards were in place.
Two applications in particular attracted my attention. The first was in the area of next-generation PON. As outlined in this issue (see "Next-Generation PON Options Promise Greater Bandwidth"), the Full Service Access Network (FSAN) group is currently figuring out what the next generation of ITU-T based PON will look like. It could be WDM-PON; it also could also beâ��or, let's be honest, will likely be — 10-Gbit/sec PON. The final standards aren't expected to be finalized for years. But that didn't stop companies from showing off WDM-PON and 10-Gbit/sec PON technology.
In some ways, the WDM-PON announcements and demonstrations are a bit more understandable than 10G PON, in that it's been indicated official WDM-PON ITU-T specifications will undoubtedly take longer to develop — so there's less reason to wait for the standards process to take its course if you think there's a near-term demand for the technology. Clearly there are differences between the WDM-PON approaches on offer (you can review these differences now by reading the announcements on our web site, www.lightwaveonline.com, or wait for our print coverage next month), so the market will likely have a chance to decide how WDM-PON will progress in advance of any standards work in the area — assuming, of course, that carriers really are interested in the technology within the next year or two.
The second application where NXTcomm activity anticipated standards activity — 100-Gbit/sec transport — illustrated a more subtle reason for companies to position themselves in the market ahead of the relevant standards activity. The 100-Gbit/sec announcements and demos at NXTcomm could roughly be divided into two areas. The first can be described as demonstrations of proprietary approaches that were unlikely to match whatever the IEEE or ITU-T are likely to codify into their standards. Like the WDM-PON activity just described, proprietary approaches promise to meet an immediate need. Carriers are left to evaluate these technologies in light of future interoperability requirements at their own risk.
The second focus of 100-Gbit/sec activity involved "proof of concept" demonstrations of some aspect of the upcoming standards. Standards bodies routinely establish foundational approaches to solving a problem; companies then take these approaches and implement them to illustrate how these concepts might be realized in the real world. The point isn't to immediately sell product. It's to position the company as a go-to source for technology once the standard has been solidified — or to influence what the standard looks like by offering the relevant working group a model to follow.
This is all well and good. However, the industry must continue to tread carefully when attempting to anticipate standards activities, both "official" and "industry de facto." For example, the Optical Internetworking Forum (OIF) has embarked on development of an implementation agreement for 100-Gbit/sec DWDM. The agreement, OIF sources tell me, will focus on dual-polarization quadrature phase-shift keying (QPSK) with coherent detection as a modulation format. "Industry consensus" indicated this was the right approach to take, the source said. However, a quick poll of exhibitors at NXTcomm indicated that such a consensus may not be as solid as the OIF believes.
Vendors constantly battle to be first to market, to demonstrate that they are more in tune with customer requirements than their competitors. But in an industry where few companies can afford major missteps, the old adage of "haste makes waste" takes on an added degree of significance.
Stephen M. Hardy
Editorial Director & Associate Publisher
Erratum: In our May Tech Trends on 10G EPON chipsets, we described features of PMC-Sierra's reference design — specifically, support of 1:128 splits and autodetection — as reflecting aspects of the anticipated 10G EPON standards. These features instead reflect specific capabilities of PMC-Sierra's offering. Also, contrary to information presented in the article, the draft phase of the upcoming 10G EPON standard is closed to changes, although changes are possible at the Work Group ballot phase.