Fiber-optic cable is used to move data by bouncing beams of light down glass fibers.Only two years ago, optical networking was a business plan buzz-phrase that had investors reaching for checkbooks. Then crisis struck during 2000, as it quickly became …
Of course, not everyone sees the light. At the annual Growth and Technology Conference held by Robertson Stephens in Tel Aviv in September, Brian Bean, head of global technology banking, was not very optimistic about the prospects for Israeli optical equipment developers. In Bean’s evaluation, “They’ll fare no worse than those worldwide – lousy.” Bean compares the current technological state of optical networking with the backward state of the semiconductor industry in the 1950s. “The optical domain has another 40 years of brilliant evolution to go,” he said.
But the 50 or so Israeli companies that are focusing their efforts on the domain are hoping to bring investors some return on their investments a little earlier than 2041. And they even have cheerleaders. Jerusalem-based Israel Seed Partners is one venture fund that has several eggs in the optical networking basket. The seventh and latest addition to its portfolio is Aelis Photonics, into which Israel Seed put $2.3 million. Aelis, which has only four employees, is building a platform for dynamic optical signal processing.
“Fiber is wonderful,” said Israel Seed’s General Partner Jon Medved. But even as he praised optical networking technology, Medved said that not all investments in the sector are the same. “People bought into the dream and lost touch with reality,” he said. “Like in the dot.com sector, there was hysteria.”
The optical hysteria was in local abundance in May 2000, when Lucent Technologies purchased Israel’s Chromatis Networks, developers of metropolitan area optical networking products, for a record-breaking $4.5 billion in stock. The promise of astronomical sums like these spawned a vast number of Israeli optics companies eager to obtain the same rewards. U.S. companies such as Cisco, Lucent and Nortel Networks spent with abandon on financing and development of new optical networks. But with so many companies jumping on board, the optical ship began to sink. Intense competition drove down the prices of services and equipment and slaughtered profit margins. And compounding the industry’s woes was the dot-com meltdown, which demonstrated that not as much bandwidth would be needed as had been predicted.
The market value of major players plummeted by 80 to 90 percent, debts began to accrue, and spending suddenly came to a halt. And in August 2001, as if to send a message to Israeli companies that all was not as it had seemed, Lucent closed Chromatis, firing all 150 employees. Headlines on the optical networking news site Lightreading.com were divided between optimistic new product announcements and gory reports of drastic staff cuts and Chapter 11 filings, including “Nortel swings axe, switches CEOs” and “BrightLink cuts to the bone.”
Not all that fiber was out of business; optical companies still have a lot of work to do. “Equipment vendors have delivered only half of what optical networking promised: abundant bandwidth,” said a report published in September by the research and consulting concern Forrester. “Now, carriers plagued by collapsing prices and rapid consolidation want vendors to finish the job by making optical networks cheaper to operate and easier to manage.”
One of the major problems with current optical networks derives from the mismatch in capacity among the three areas of the market: backbone, metro and access. Backbone is the long-haul inter-city link; metro is the regional ring that connects carrier access points, and access is the first-mile connection from homes and offices. Estimates put the amount of optical fiber laid worldwide in the past two years at about 160 million kilometers. But this optical fiber was laid primarily in the backbone, and in that segment of the market, prices fell by as much as 75 percent. By contrast, because of the high costs of laying fiber in urban areas, only about 10 percent of U.S. office buildings now have fiber connections.
The result is a bandwidth choke. The backbone can currently deliver 10 gigabits per second (Gbps), and today’s in-house networks (LANs) can typically handle from 100 Mbps (a Mbps is one-thousandth of a gigabit) to one Gbps. But the whole network is let down by the middlemen: the access and metro lanes are stuck somewhere around 45 Mbps. According to the Forrester report, carriers are demanding the metro and access networks be brought up to speed. Furthermore, carriers want technology with a smaller footprint that makes fibers more cost-effective and less power-hungry.
Not surprisingly, this is where many of the Israeli companies are focusing their energies. Most are still at the pre-sales stage, conducting trials and aiming to ship products by the end of 2001 or the beginning of 2002.
For example, the 3-year-old LaserComm, co-founded by Yochay Danziger, a graduate of the Weizmann Institute of Science, is combating chromatic dispersion of light in optical networks, a phenomenon that hampers higher bandwidth, the number of channels per fiber and the distances covered. The company’s “box” is slipped into the existing amplification systems, which sit every 100 kilometers or so along the network. The technology is in trials with nine of the top 10 system developers, said LaserComm Managing Director Mickey Steiner, and it works with every type of fiber.
Steiner runs the company’s development center in Tel Aviv, but the manufacturing is done in Plano, Texas, close to the major customers. Steiner disagrees with the assessment that the industry is a long way from maturity. “I don’t think it is going to take 40 years, or even 14 years, to mature,” he said. “Today the demand is there and growing tremendously. All-optical networks are something that we feel will develop very fast in the next few years.” Unlike current optical networks, which still use electronic components for tasks such as switching and routing, all-optical networks do not use any electronic components.