Optoelectronics industry begins restructuring

Jan. 1, 2003

The dramatic drop in sales revenue over the past two years has driven drastic restructuring of the major vendors at the network and component levels of the fiber-optic industry. As growth resumes in 2003-04, the industry structure will have changed drastically from 1999-2000.

Digital transmission is growing much faster than analog. Thus, digital traffic has advanced from less than 20% of throughput in 1990 to more than 50% in 2000 and is projected to reach beyond 90% of rate times distance loading by 2010.

Economics, plus the concern of the U.S. Department of Defense about the security of the existing (SONET) telecommunications network, fueled the expansion of the Internet and its supporting protocol, equipment, and components. ElectroniCast designates this Internet-related market as "digital-centric," in contrast to the traditional analog-centric telecommunications network. Thus, the total "wireline" communications market is bimodal, with significant differences at all levels (see Figure). It is the digital-centric side of this structure that drove the most dynamic 1997-2000 growth and suffered the most dramatic 2000-02 deflation.

Digital-centric optical links range from global-network long-haul (LH) to millimeter interconnect of optoelectronic ICs. They can be segmented according to application: on-chip (millimeter), intra-equipment, premises/LAN, MAN/WAN (access/metro), and LH. While build-out of LH digital-centric links drastically reduced in 2001-02 and other applications have declined as well, development and production at the millimeter-to-10-m level expanded in late 2002. This activity supports next-generation servers, routers, and optical (OEO) crossconnect switches. SAN deployment and SAN-to-SAN interconnect is a strong demand driver for such equipment.

The causes of the optical industry downturn have been widely discussed. However, the initial 2000 collapse of digital-centric network expansion funding was exacerbated by a series of confidence-deflating factors:

  • The 9/11 attacks and consequent major funding of counter-terrorist measures.
  • The "Enron bomb"—the perception that numerous leading high-tech ventures had published grossly misleading financial reports and that several "independent" accounting firms appeared to be complicit.
  • The U.S. government's moves toward major unilateral armed intervention into other countries, raising a perception of major investment in military resources.

Given these successive blows, it is encouraging that the investment community, as 2002 gives way to 2003, is funding a semiconductor industry recovery and actively reviewing fiber optics-related opportunities.

Throughout 2001-02, the consumption of communications services has continued its impressive growth. Despite overall capital-expenditure reductions at many U.S. service providers during 2002, analog-centric regulated telcos globally have continually increased their spending for capacity expansion, especially in the metro/access space.

As mentioned, there has been continuing rapid growth in optical interconnects at the low-data-rate end of the network. Conversely, four-channel fiber links at 400 Gbits/sec per channel will become more common; some links will soon reach 1.6 Tbits/sec. There is huge pent-up demand for 2-Mbit/sec Internet connections, and HDTV finally is building customer momentum.

At the high-volume subscriber connections, fiber must compete with copper. Thus, the fiber components market is highly price-elastic; large drops in price will be more than offset by increased consumed quantities. The optical components industry increasingly will resemble the semiconductor industry. With high-end components at 10 and 40 Gbits/sec, compound semiconductors will play an expanding role. The premises, MAN, and LH networks will function as computer networks.

This trend toward digital-centric communications, with an emphasis on high-volume/low-cost semiconductor-based components, has been noticed by semiconductor firms. Intel has invested an estimated $1 billion over the past five years to reconfigure itself to include a strong optical interconnect leg. TriQuint Semiconductor has acquired the extensive optoelectronic-component line of Agere Systems, and Bookham Technology similarly has folded in Nortel Networks' optical components. Vitesse has leveraged its gallium arsenide and indium phosphide base in off-the-shelf ASICs into optical transmitters/receivers. Pioneer Infineon Technologies, with alternate-source Molex, leads the surge into multifiber/multichannel low-cost/short-reach links.

This rapidly developing hardware-based trend by major corporations into high-volume, low-cost optoelectronics will spark the attention of investors and lead to a movement back to digital-centric deployment support in 2003.

Jeff D. Montgomery is chairman and founder of ElectroniCast Corp. (San Mateo, CA), an optical technology market research and analysis firm.