The emergence of 10-Gigabit Ethernet networks has fostered the development of new optical fibers and optical-fiber cables. These new fiber cables—especially the new laser-optimized 50-µm multimode cable that supports 10-Gigabit Ethernet for distances of 300 meters using inexpensive vertical cavity surface-emitting lasers (VCSELs)—are expected to expand the market for new applications.
Currently, the most popular application for optical-fiber cabling is the system riser or backbone. In the future, I anticipate fiber-to-the-desk as a significant new horizontal application for optical-fiber cabling. These two, however, are not the only applications for fiber cabling inside buildings. Three relatively new applications, all of which require gigabit speeds, are emerging: storage area networks (SANs), switch-to-switch interconnectivity, and fiber-to-the-zone.
Of these horizontal applications, currently fiber-to-the-zone is the largest. Over the next five years, I expect fiber-to-the-desk to become the largest horizontal application, with the highest growth rate. But over the forecast period that ends in 2007, riser fiber applications will remain the largest in-building application for fiber cabling—larger than the total of all four horizontal applications.
The new 50-µm laser-optimized multimode fiber, which supports 10-Gigabit Ethernet, will be a primary cable of choice over the next five years. These fiber cables are ideal for switch-to-switch interconnectivity and SAN applications. I anticipate that in the future, fiber producers will develop 50-µm technology that will support 100-Gbit/sec speeds.
Additionally, I believe that enhanced fiber cables targeting fiber-to-the-desk applications will be developed. These cables will provide longer distance capabilities, supporting a centralized all-fiber architecture that incorporates physical-hardened properties as fiber moves closer to workstations.
But I am concerned about confusion in the marketplace. Consider, for example, that specifiers and users have the opportunity (or burden) to choose from "enhanced," "laser-optimized," and "original" 50-µm multimode fiber. In my recent study, I categorized all the current and expected future fiber types by performance levels—from 10 Mbits/sec to 100 Gbits/sec—and came up with seven grades of cables, a subset of which is appropriate for fiber-to-the-desk applications. In the study, I describe the grades and make fiber-year forecasts for each of them.
Over the next five years, the Gigabit Ethernet cabling market is projected to grow at 29%, based on total copper and fiber shipment values. That's a significantly higher rate than the total building-cabling market growth of 6%. Unshielded twisted-pair copper cabling—specifically Category 6 cabling—will be the primary cabling used for the next several years to support 1-Gbit/sec applications. Beyond 1 Gigabit, fiber cabling is required at distances to 100 meters. As a result, I expect that by the 2004 time frame, fiber will begin displacing copper cabling for the higher-speed requirements. I project a 31.9% growth for fiber cabling, with fiber-to-the-desk applications leading this growth.
By 2007, Gigabit Ethernet fiber applications will account for 70.7% of the building fiber-cabling market. One apparent trend is that the current limitation of 10 Gbits/sec requires a next-generation, 100-Gbit/ sec protocol to foster its growth. When the time comes for users to consider 10 Gbits/ sec in a horizontal application, they will be reluctant to do so if the backbone system creates a bottleneck.
Thus, with the emergence of a 100-Gbit/sec system expected around 2005, the 10-Gbit/sec cabling market is expected to expand its growth during that time.
