Active Optical Cables Market Analysis 2009

Abstract

“Active Optical Cables Market Analysis 2009” examines several focus segments generating significant business growth over the next five years. Fifteen separate market segment forecasts, covering four platform types, four bus standards, and five platform categories over a five-year forward time-interval, were analyzed as part of the overall Active Optical Cable (AOC) Market Analysis 2009.

Overall cumulative cable revenue is expected to exceed $8.5 billion (US) from 2009 through 2013 just for the application segments included in this report. Single-year revenue is expected to grow from the present-year level of <$100 million to over $1.1 billion in 2010, then up to over $2.6 billion by 2013. As additional application segments are included in the overall Market Analysis, the total revenue numbers should be expected to increase substantially over these conservative figures.

Total AOC cable count identified in this report should exceed 2 million units in 2009, and grow to over 75 million units by 2013. As in the revenue analysis above, we expect these AOC figures to expand as more application segments are studied, and as better attach rate forecasts become available. Growth in the overall AOC forecast should generate better supplier cost efficiency at both the AOC OEM cable providers as well as with their underlying material supplier communities.

Certain market forces influence changes in AOC attach rates used in this analysis. Several AOC users reported shifting away from copper cables towards AOC alternatives, even when the copper cables were technically cheaper for their shorter cable length application usage. One notable case was IBM' s Roadrunner supercomputer, which reportedly switched entirely to AOC, reporting 55 miles of cables in their first installation. Initial expectations were to use AOC models only for cables 5-10 meters or longer. As more and more customers experience this “Waterfall Effect,” significant increases in the overall AOC attach rate should occur, potentially 2-3x over the volumes included in this report.

Annual fiber shipments for active optical cables is expected to grow from ~13 million meters in 2009 to almost 434 million meters by 2013. Five-year cumulative shipments should total 1040 million meters. Certain OEMs use different designs for cables under 20 meters versus those in the 20-100 meter range. OEMs that target common optoelectronics designs capable of supporting broader cable length ranges will benefit from lower overall product costs as well as increased market response flexibility to changes in customer cable length preferences over time.

The average cable length covered by this study ranges from 1 to 100 meters. Volume is limited at the low end (<5 meters) by cost-efficient copper cable equivalents. At the 100-meter and longer range, more-effective passive optical designs take precedence with the customer base. In some application segments (such as mainframes), we expect to see an appreciable volume growth in the &ldt;5-meter cable length range, especially as supercomputers move towards quad-data-rate (40Gbps) fabrics. We anticipate a similar optical cable preference improvement in consumer applications as their preferred bus speeds begin to advance beyond 10Gbps .

As expected, the bulk of the AOC business is centered on models in the 10-20-Gigabits-per-second (Gbps) range. For cables carrying digital data below 5Gbps, including those <5 meters in length, copper can still provide suitable cost-effective technical solutions. For cables carrying data greater than 40Gbps, AOC usage can be expected to dominate copper cable share. While the number of 40Gbps ports today is relatively small compared to other speed groups, there is substantially growth in this segment, with its higher average attach rates making it a very attractive segment for AOC providers and their core materials providers.

AOC cable OEMs, as well as their underlying material suppliers (i.e., fiber-optic cable, VCSELs, optical detector/PIN diodes, and connectors), should benefit substantially from this growth, especially those with earlier market entry points.

Report Features

Several key high-volume applications segments were chosen for this report. Emphasis was given to ones with products using digital electronics interfaces running either at or above 5Gbps, as well as those demanding extended cable lengths beyond 2-5 meters at these speeds. It is already well understood that copper-based cables are experiencing significant challenges in this performance range, thus offering significant copper displacement sales opportunity for AOC providers.

• Mainframe/Supercomputer (HPCC)
• Desktop and Notebook/Portable Personal Computers (PC)
• High-definition Television (HDTV)
• Consumer Electronics devices (CE)

We chose specific sections of these markets where we felt there was reliable information capable of creating an AOC demand forecast in these areas. Other application segments can certainly be considered as part of future versions of this report, or as part of custom analysis support.

• InfiniBand
• USB
• HDMI
• DisplayPort

All of these interfaces are either operating at or above 5Gbps today, or expected to be at this level within one to two years. This report includes segment breakdowns by bus speed, as all of these interfaces are expected to ship two or more speed versions over the five-year interval covered by this report.

While there are other bus interfaces that could meet the speed and/or distance requirements, we chose not to include that information at this time, as we felt the association with AOC demand was not currently strong enough to warrant inclusion in this report. In some cases, viable passive optical interconnects were already in place, making AOC solutions less attractive.

Forecast information was generated from studies on the number of new ports of each interface type expected over the study interval. Since copper cables are expected to still be used by many users of these new systems, we used fairly conservative AOC attach rates to create the actual AOC forecast data. Actual market demand could be significantly higher than the conservative data used in this report as customers realize additional benefits from optical cabling solutions.

• Revenue
• Number of cables sold
• Expected length of optical fiber and cable jacketing
• Number of connectors, including segmentation by application and interface type
• VCSEL counts, including breakout by VCSEL speeds

Individual tables and graphs provide the reader with overall AOC figures, as well as breakouts by application segment, product categories, and bus interfaces.

Other possible growth factors should be considered when weighing the forecast figures in this report. Each of the above application sectors and bus interface types have substantial installed base of pre-existing units. As consumers elect to increase their cable reach, they are likely to consider optical fiber cables such as AOC, instead of simply choosing longer versions of their existing copper cables. As AOC costs come down over time, and as consumers become more familiar with optical cables, they are more likely to choose AOC alternatives.

In many cases, the above segments have an installed base far larger than the new port creation figures mentioned in the analysis. If replacement cable rates with AOC were to follow the trend for new port creation, the overall AOC volumes would be substantially larger than the figures included as part of this study. While we are tracking the installed base and new port creation rates for the named segments in this report, we chose to exclude this business upside factor in the overall analysis, at least until such time as we get more reliable information on actual attach rates against the actual installed base.

Simple online searches point to several cable providers and test labs (such as, but not limited to, MonsterCable and SimplayHD) actively promoting educational programs to raise consumer awareness of copper cable performance in these types of higher-speed applications. At the same time, consumers are becoming more aware of AOC cable performance relative to even the best copper cables, thus creating a corresponding increase in AOC attach rates. Once more, we did not find hard quantitative data to include in the actual figures shown in this version of the study, but will be considering this factor for future studies.