New Markets for Telecom and Datacom Lasers: 2007 to 2012

Disclaimer: Plese note that the pdf version a 1-5 user license

With the renaissance of the optical networking business, the volume demand for lasers has increased considerably. This is good news for laser makers, of course, but unfortunately for them lasers have increasingly become a commodity item. This is true even for high spec telecom lasers to some degree. The bottom line - literally - is that laser firms must either learn to make money with low and competitively vulnerable margins or they must find some way to innovate technologically in way that radically shifts price points for a given laser performance.

Packaging, Optical Integration and Silicon Photonics

Two areas where such innovation seems most likely to occur are in packaging and integration. Packaging is a large part of the cost of lasers and so cost changes there can make a real difference. Integration brings with it - at least potentially - all the benefits that have driven the semiconductor market in the past.

Companies already build integrated devices with lasers and monitoring diodes and EDC. The opportunities for laser manufacturers are to either partner with a company doing this work or develop integration techniques of their own. This is the wave of the future and holds much promise for cost reducing optical components and packaging.

Meanwhile, silicon photonics technology appears very close to breaking through into the commercial marketplace and has the backing of some big name firms, such as Intel, as well as smaller innovative one, such as Luxtera. It promises long-wavelength transceivers on cost parity with 850 nm optics. CMOS photonics could be the single most important technology that finally pushes short-wave optics and MMF into oblivion. With silicon photonics there is the promise of a 1310 nm or 1550 nm device that can be purchased for just over two times the cost of an 850 nm one, but go more than 10 times the distance. Whether this promise will be fulfilled is still an open question.

The Tunable Transition

Another area where laser firms may be able to squeeze more profitability from the laser segment will be in the transition to tunable lasers. While FP and DFB lasers are still very much the norm, the shift towards WDM in the metro, means more tunables a technology that is still not mature.

How quickly this will happen all depends on how rapidly tunable laser suppliers can cost reduce their products. The dominant technology for tunable lasers has been to use an array of DFBs, but common sense suggests that an integrated DBR solution will be cheaper. But it is not yet clear that a DBR strategy could be implemented for high volumes of lasers. Currently, equipment vendors and service providers seem content on a fairly slow adoption for tunable lasers. However, this may change if an innovative solution to cost reduction can be found - perhaps a monolithically integrated DBR solution with a wavelength stabilizer for long-term reliability?

FTTx: A Laser in Every Home?

Integration and packaging are areas of innovation in the FTTx sector as they are elsewhere, but this is one area where potentially laser suppliers could actually build a business based on paper thin margins and very large volumes; a similar business in fact to the CD/DVD laser business.

Verizons FiOS rollout continues to fuel the need for diplexers and triplexers for its PON networks in the U.S. and there are many opportunities in Europe and Asia as well. Asia, in particular Japan, is the fastest growing market for FTTx with its implementation of EPON. South Korean networks are also an important source of demand for FTTH lasers. But while the ""laser in every home"" aspect of FTTx is something it shares with the CD/DVD laser business model, the FTTH laser business is ultimately controlled by a handful of consumers such as NTT and Verizon who can drive a very hard bargain with regard to price. Some suppliers of diplexers and triplexers we have talked with have found that the FTTx can be - at best - barely profitable.

FTTx may currently a growing opportunity for laser suppliers, however, but there are still too many such suppliers in the market today. Consolidations seem inevitable, but those companies that survive will then be able to count on large volume prospects all over the world.

Opportunities of the Datacom Kind

Datacom lasers may be a better bet - from the perspective of laser suppliers - in terms of oligopolist customers controlling price points, although Cisco most certainly does have this power to some extent. And datacom lasers are even further along the path to technological maturity than FTTH lasers; more than 85-percent of todays Ethernet systems use 850-nm optics.

What should give laser makers some hope for the future, however, is that data rates are increasing. 10 Gbps is mainstream in many applications and we believe it will follow its ancestor Ethernets to ubiquity. 100 GigE, a standard that no one would have taken seriously five years ago, is now being taken very seriously indeed.

40 Gbps technology has been around for more than seven years, but has just recently been deployed in the public network - AT&T, Sprint and Softbank have it in their backbone networks. HPC trial networks are the only ones running above 40-Gig right now. One of the big open issues for 40G and beyond is the modulators. Several advanced modulation schemes will be adopted for long-haul 40-Gigabit transmission. For metro, access and enterprise networks, most equipment manufacturers and service providers seem to be leaning towards multiple channels of 10-Gigabit versus serial data rates of 40 or 100 Gbps. This leaves many opportunities for laser manufacturers for both short and long-reach applications. In fact, Infinera is already using a 10x10G array in its equipment and has demonstrated 40x40G arrays. Avago Technologies also demonstrated 10x10G long-wavelength transmitter and receiver modules at this years OFC.

Since the technology already exists, it is just a matter of determining the packaging and making it production-worthy.

The Future: Is this the End of 850 nm?

All this speaks to opportunities for laser makers in the future. Probably many of these opportunities will not fully present themselves until after the timeframe considered in the forecasts in this report, but they will certainly occur in the next 10 years or so. Unless the problems associated with high current density and overheating at 10-Gbps speeds and above are solved - an opportunity in itself -- the next generation of lasers will become long wavelength ones only. What will accelerate the 850-nm laser vanishing act will be the onset of technology that allows long-wavelength devices to become cost effective compared to short-wavelength versions. Recent developments in silicon photonics, integrated optics and the packaging of transceivers show the accelerating trend towards enabling this transition. For these higher speed lasers there is plenty of opportunity in how they will be modulated, amplified, cooled and dispersion compensated.

Even though the laser business is very competitive and somewhat commodity, there are still plenty of opportunities for those companies willing to stay the course.

Goals and Scope of this Report

The key objective of the report is to determine and quantify where the major opportunities are to be found in the telecom and datacom laser business over the next five years.

The report covers the full range of laser architecture and technologies (DFB, FP, DBR, VCSEL, tunables, etc.) and materials platforms. We also examine the laser requirements, market drivers and challenges for each of the main application areas; acess/FTTH, metro, long-haul, LAN and SAN. The scope of the report is universal in terms of data rates supported. However, because the report is primarily concerned with new opportunities, our focus is on 10 Gbps networks, which are here now and a not completely commoditized as far as lasers are concerned. It is also on emerging 40, 100 and 160 Gbps networks.

Since we believe that significant cost improvements may result from innovations in packaging and integration, these are also topics that we return to throughout the report. While the forecasts at the end of the report are for five years only, we comment on likely developments in the laser space beyond that timeframe, which is when the transition to much higher speeds will have major impacts in terms of new business opportunities. Almost every firm in the laser space - and beyond that almost every firm in the entire telecom/datacom components space - are now becoming fixated on the (possibly) glittering prizes at 40 Gbps and 100 Gbps.

We have also included in this report an analysis of the strategies being adopted by laser firms both large and small and, as with all CIR reports, this study also includes a detailed five-year forecast of the markets covered.

Methodology of this Report

The information and analysis contained in this report are based on an extensive interview program conducted primarily during the second and third quarters of 2007. Firms covered in this program included most of the largest suppliers of laser devices, as well as important start ups. We have also based our analysis on the views of carriers and network equipment firms. While no claim is made that our selection of firms for interviewing is ""scientific"" in any precise sense, we did try to include firms that were either influential in the segments of the markets that they served or could potentially be so. In our interviews, we did not use a structured questionnaire, but instead focused our discussions with the firms interviewed on their particular area of activity. This approach, CIR believes, leads to the deepest insight into what is really driving markets.

In addition to the primary research carried out for this report, CIR also conducted an extensive search of secondary sources including research publications, white papers, corporate web sites, SEC information, conference proceedings and previous CIR reports.

The forecasting methodology used in this report is discussed in greater detail in chapter four -- the chapter of the report that deals with forecasting. However, it is based on the modeling techniques that have been developed by CIR over the past 25 years. The actual models used have, however, been updated to reflect the radically new situation that we find in the optical components business at the present time.

Plan of Report

Chapter Two of this report analyzes discusses the technological evolution of telecom and datacom lasers. In Chapter Three, we look at the market requirements for these devices. Finally in Chapter Four we forecast the market for each of the major communications laser market segments.