LanOptics is dead, long live EZchip

LanOptics chief Eli Fruchter on where he and network processors are headed.

"We keep asking ourselves where we can grow," says Eli Fruchter, founder and CEO of network processor LanOptics (Nasdaq: EZCH). "Firstly, we aim to be the biggest company in the market for network processors. True, this is not a big market, but we want to be the 'Intel' of network processors."

Today, the company's shareholders meeting is due to give LanOptics the green light to change its name that of its subsidiary, EZchip Technologies Ltd.. This will complete the companies' merger, after LanOptics announced in January that the remaining funds with shares in EZchip had agreed to exchange them for shares in its parent, LanOptics. The company, which has a market cap of $300 million on Nasdaq and the Tel Aviv Stock Exchange (TASE), also changed its symbol in January from LNOP to EZCH.

"For me, the name change brings me full circle," says Fruchter. "I founded EZchip in 1999 as a spin-off from LanOptics, and today all our products are sold under the EZchip brand." Fruchter founded LanOptics itself in 1990, for the purpose of creating telecommunications solutions for "Token Ring", a communications protocol invented and promoted by IBM Corp. (NYSE: IBM). The start was promising - LanOptics products soon began selling, and within just over two years, the company had floated on Nasdaq at a value of $60 million.

LanOptics' sales for the Token Ring environment continued to grow, until the arrival of new communications protocol, Ethernet, which ultimately caught on worldwide.

"In 1997-98, I switched the company to Ethernet," Fruchter continues, "but here we weren't the first as we were with Token Ring. We waited to see whether there was something unique that we could do. In 1999, we saw that there was a niche for network processors, and then I founded EZchip to make high-speed Ethernet network processors. At the time, LanOptics was still selling boxes to EZchip's competitors in the network processor field, and we couldn't enter that business under the name LanOptics."

LanOptics continued to sell its boxes, which sat in computer rooms at large enterprises, but all its R&D budget was transferred to the new subsidiary. To finance the development, LanOptics looked for partners that would invest in its subsidiary, and found them in the venture capital funds which invested $60 million in EZchip over several rounds.

EZchip began developing high-speed network processors, which sit on routers in large Internet networks. Every such router, with EZchip's processors attached to it, provides Internet access to an entire neighborhood or city, as opposed to low-speed network processors, which sit on small routers which serve a single building or street.

There are actually three key customers for high-speed processors, which share 90% of the market between them - Cisco Systems Inc. (Nasdaq: CSCO) which has 50%, and Alcatel-Lucent (NYSE: ALU) and Juniper Networks (Nasdaq: JNPR), with 20% each. Two of the three are EZchip customers, with past reports naming Cisco and Juniper, while Alcatel-Lucent continues to develop its own processors for the routers it sells.

Gross margins - the next generation

"Our competition is not just with other companies which develop network processors," explains Fruchter. "The really stiff competition is internal - with R&D departments at equipment producers. We have to reach a state where it won't pay equipment producers to develop processors in-house, and that it won't be worth their while committing themselves to a development process which could take five years and cost more than $100 million."

In 2002, EZchip completed work on the prototype of its first product, the NP-1, and it has earned revenue since 2003. The next product, NP-2, went on sale in 2006, and now accounts for 80% of the company's sales.

EZchip developed the prototype for the third generation of the processor, NP-3 together with semiconductor giant Marvell Technology Group (Nasdaq: MRVL), for a leading customer in the sector. The first orders from the customer will be delivered later this year, something that should increase EZchip's revenue and also its gross margins. "Marvell will make the sales, and pay us royalties, which will be as high as the profit we would have made had we produced and sold the processor on our own," explains Fruchter. "Every processor sold through Marvell will yield a gross margin of 100%."

EZchip believes that, at the end of the year, with the move to mass production for the large customer, the company's profit margins will rise from the current 60% to 70%. "We also have a second model of the NP-3 for the rest of the market, and we've only just received samples of it for testing. I expect it to start selling by early next year," Fruchter says.

"The cycles in this market are very long. If we are awarded a design win, it will take us two years to develop the product, and then another year for it to go through the entire testing process at Internet operators and start generating revenue. Once this is complete, it will sell well for a few years, during which we bring out the next product, and look for a design win for the generation that is next in line," says Fruchter. "We’re constantly in this kind of situation - selling an existing product and developing two generations forward."

Low speeds as well

At the beginning of 2008, EZchip announced it had obtained the first design wins for the next generation of its processors - NP-4. "Aside from the fact that NP-4 will be a lot faster and more powerful, it will also come with video processing capabilities," says Fruchter. "There are forecasts which say that in 2011, half of the traffic over the network will be video. The NP-4 has been built to provide solutions for this."

The long development cycles, and the collaboration with the giant companies which produce the telecommunications infrastructure that powers the Internet, enables EZchip to envisage how the communications world will look several years hence. "Technically speaking, the Internet still does not have the bandwidth to deliver large quantities of video to every home," says Fruchter. "But in your home, in a few years from now, instead of a cable socket you will have an Internet socket, and you will either you connect your television to it and see all the channels, or your computer and see everything that you see on the television screen. Some people even believe that the television will disappear altogether."

Fruchter explains, "Today, when you open a video file on a computer, it takes ages for it to appear, and sometimes it stops and then carries on loading. If that were to happen while switching channels on the television, we wouldn't stand for it. The professional term is 'quality of user experience'. The quality of the experience on the Internet should be no less than on cable, and that is how it will be," he promises.

A few months ago, EZchip announced it would be entering low-speed processor production, with a new family of products called NPA. "This chip is a lot smaller and much cheaper, and the development cycles are shorter," Fruchter concludes.

Published by Globes [online], Israel business news - www.globes-online.com - on July 10, 2008

© Copyright of Globes Publisher Itonut (1983) Ltd. 2008

Twitter Facebook Linkedin RSS Newsletters גלובס Israel Business Conference 2018