End-to-End Optical Ethernet a “Dream” – Report

It hasn’t been that long since the industry started referring to “Optical Ethernet” — previously, it was called 10/100 Mbps, 10BASE-T, 100BASE-T, 1000BASE-T or Gigabit Ethernet. Even as the new term becomes de rigueur in the optical-networking industry, a new study from Communications Industry Researchers Inc. (CIR) shows that the majority of the Ethernet ports in the near future will remain copper based.

In a white paper released prior to a report called “Optical Ethernet: A Market Analysis of Gigabit and 10-Gigabit Ethernet Transceivers”, the research firm said that there is a market for Optical Ethernet and the optical transceivers to address the technology, but that market is “much slower growing” than originally anticipated and will remain so in the near term.

CIR also said it does not see widespread end-to-end Optical Ethernet systems emerging in the near future, if ever. While Fiber To The Building (FTTB) applications will continue to creep upward, Fiber To The Home (FTTH) might be decades away from becoming big, much less fiber to the desktop. For a long time, a common scenario will be a combination of electronic Ethernet in the LAN with Optical Ethernet in the access network and MAN.

Today and for some time to come, the largest market for Ethernet is in the local-area network (LAN). Enterprises using 10/100 Mbps Ethernet connections comprise most of the sales for leading equipment manufacturers 3COM, Cisco, Enterasys, Extreme Networks and Foundry Networks. Gigabit Ethernet is starting to make inroads in large data-centers at the edge of the public network, in the access network with competitive carriers such as Yipes and Cogent and also in the metropolitan space in the case of Telseon.

Most Gigabit Ethernet optical transceivers will be used in the access and metro markets and not in the enterprise market, CIR also asserted. They will be used primarily in uplink or aggregation ports for 10/100 connections from LANs. Optical transceivers may also be utilized in Virtual Private Networks (VPNs), as they connect companies that have several sites within a metropolitan area. Whether the access and metro solutions will predominately be straight Ethernet or Ethernet over SONET has yet to be determined.

CIR Senior Optical Components Analyst Lisa Huff wrote in the report that the Optical Ethernet Transceiver market will grow from almost $600 million in 2002 to just shy of $1.2 billion in 2006. Ten-Gbps Ethernet will have limited potential, although 1-Gbps Ethernet offers significant opportunities, she also said. But CIR believes that the lion’s share of this growth will be solely for Gigabit Ethernet modules.

For Optical Ethernet to succeed, Huff said the market will need to meet major factors:

  • High level of vertical integration. It is key to control costs when entering such a price sensitive market as Ethernet. Therefore, owning intellectual property as well as having as much control as possible over its supply of components, is a critical key to success.
  • Financial stability. Optical Ethernet companies need large cash reserves to get them through leaner times and to deal with the loss of a major customer, because some equipment vendors have had no qualms about abandoning transceiver suppliers, especially in order to get significantly lower prices.
  • Quick time to market on new designs. While profits in the Ethernet transceiver business are relatively low, new products/form factors in this space command somewhat higher margins. Because this market is something of a cutthroat one, those, which are first into the market with new designs, are the only players that ever really take much advantage of bigger margins.
  • High volume manufacturing and sales . Ethernet equipment vendors place a lot of pressure on the highest cost component of their systems — optical transceivers. Success in the Optical Ethernet market requires transceiver vendors to design their sales and manufacturing operations for high volumes to make up for the generally low margins that this business tends to offer.
  • Large annual price reductions. Our research indicates that Optical Ethernet transceiver vendors will need to keep making 10 to 20 percent price reductions per year to maintain a high share of the market.

Vendors, especially newcomers, can gain a major edge over market leaders Agilent and Finisar by moving to small form-factor pluggables (SPFs), extending long-wave VSCELs to longer wavelengths, and developing into large 10-Gigabit laser-chip makers.

For companies, Optical Ethernet is a good value proposition that “appears to be very in tune with the current cost-sensitive nature of the optical networking business,” Huff said. For some components companies, at least, Optical Ethernet may be a much more attractive new business option compared to optical switches or integrated optical chips, which may have huge long-term potential but are likely to produce negligible revenues in the next two years, she added.

The full report will be published in November by CIR.

Bob Woods is the managing editor of OpticallyNetworked.com.

News Around the Web