by Stephen Hardy
The year may be almost a quarter old, but it's still young enough that component suppliers continue to test the assumptions on which they based this year's plans. Two of the larger optical technology suppliers, JDSU (www.jdsu.com) and Bookham (www.bookham.com), have based their strategies at least in part on the expectation that the trend toward reconfigurability will continue to require new component and subsystem advances, particularly in the area of tunability. Functional integration, particularly at the wafer level, will prove essential, sources at these two firms predict. Product line integration, in which the companies offer a range of components and subsystems for a particular application, also is a strategy JDSU and Bookham share. However, their takes on how to meet customer requirements don't agree completely.
But let's start where the companies' visions mesh. "Network agility – I think for the most part in 2007, it kind of became table stakes," offers Craig Iwata, senior director of marketing and business operations, optical communications, JDSU. While much of the talk about adding flexibility through the deployment of reconfigurable optical add/drop multiplexers (ROADMs) in 2006 was just that – talk – carriers put their words into action last year, Iwata asserts. For JDSU, this meant increased sales for its ROADM subsystems, particularly its planar lightwave circuit (PLC)-based wavelength-selective switches (WSSs). Its tunable laser and transceiver products also benefited from the overall desire for flexibility, Iwata adds.
Bookham product and sales management also has moved to take advantage of this trend, particularly with its tunable laser portfolio. "I really think that Bookham is perfectly positioned to be the dominant player in that reconfigurable/tunable market," asserts Yves LeMaitre, who recently left Avanex to become Bookham's vice president of global sales for the telecommunications division and vice president of corporate marketing. "There's competition – there are good players, solid companies. But if you look at the fundamental strength in technologies that exists at Bookham, I really feel that the company is very well positioned."
That said, LeMaitre concedes Bookham's product line for reconfigurable applications remains incomplete. "I think a WSS is one that is obvious to everyone," LeMaitre concedes in terms of offerings his company lacks.
"It's an area where we need to play; it's an area where our customers want us to play," adds Adrian Meldrum, Bookham executive vice president and general manager of the telecommunications division. "It's not an area where we have in-house technology or core technology to address that. So we continue to evaluate how best to play in that space in a cost-effective way with potential other partners in that field."
Meanwhile, both companies plan to focus this year on the development of pluggable transceivers that are also tunable. JDSU recently announced what Iwata considers a major step in this direction, a tunable transmit optical subassembly (TOSA) based on the company's integrated laser/Mach Zehnder (ILMZ) device announced last September. The integration of the laser and modulator onto the same chip opens the door to package sizes that will enable the delivery of tunable transceivers in form factors as small as the XFP. Iwata says that JDSU hasn't decided whether it will make the ILMZ available to other companies, but has every intention of offering a JDSU-branded tunable transceiver based on the chip.
With the tunable TOSA in hand, Iwata says that emerging module form factors such as the XFP-E (see "XFP-E Compromises Size for Pluggability, Tunability" by Meghan Fuller Hanna, Lightwave, July 2006, p. 11) – which is essentially double the width of an XFP – won't be necessary. "We don't have customers coming to us and asking us for a double-wide format," he explains. "We believe that people are looking at the double-wide form factor because they don't believe they can fit tunable into an XFP form factor. We believe we can fit it into an XFP form factor."
Bookham, which also has ILMZ technology, does see promise in the XFP-E. "There seems to be wide agreement that the XFP-E format is probably the right format for the play," Meldrum counters. "But having said that, there will be other formats going to market before that in specific cases."
In addition to lasers, tunability makes sense for dispersion compensation, the JDSU and Bookham sources agree. Neither company has a standalone product in this area, however. Iwata points out that tunable dispersion compensation will prove particularly important with the advent of transmission rates of 40 Gbit/sec and above –and the 40-Gbit/sec transponder that JDSU co-developed with Mintera (www.mintera.com) contains an integrated tunable optical dispersion compensator.
LeMaitre notes that dispersion compensation can take electronic as well as optical forms. Bookham hasn't publicly announced its strategy in the tunable dispersion area.
Iwata believes JDSU's tunable TOSA highlights both a skill set and an overall trend in the optical components and subsystems space: functional integration. "To us, functional integration is combining different functionalities in a way where we can decrease the size of something or decrease the cost of something, increase the power efficiency, and increase the performance," he explains, adding that the best integration strategies achieve several of these goals in combination.
The company's ROADM offerings exhibit an even better example of the benefits – and the scale – of such integration efforts. Iwata says that JDSU's most recent generation of PLC-based ROADMs integrates the functions normally provided by 491 discrete components into four chips. Much of this kind of integration happens at the wafer level, and Iwata believes that JDSU is in a good position to continue to advance on this front because of its in-house fabrication and wafer design expertise.
Bookham also has its own wafer fabs that it expects to leverage for wafer-scale integration. In addition to applications such as the ILMZ, Meldrum believes chip-level integration will offer avenues for cost reduction in 40-Gbit/sec transmission. "One of the problems in the space at the moment is that there are people doing 40 gig, there are people deploying 40 gig, using our tunable lasers in many instances, but understanding how people can do that in a cost-effective way and actually make money at it is not something that's clear," Meldrum explains. "But that's where we believe our indium phosphide platform really comes in, in terms of the integration, keeping the complexity at the chip level and therefore coming up with a cost-viable play."
But functional integration can happen at many levels. Iwata points to the opportunities available within circuit packs, where the decreasing size of individual modules can enable subsystem suppliers to cram more functions into the same space.
Meanwhile, product line integration offers suppliers a way to leverage their expertise across multiple markets. Meldrum points to Bookham's activities in the tunable space – in which it offers tunable lasers, tunable transmitter assemblies, and has just announced a tunable small-form-factor 300-pin transponder –as an example.
"It's coupling that technology with the vertical integration that comes from playing at all of those different levels, both on the product side and the manufacturing side," he concludes.