Copper Access Startups Tarnished
The developments are disconcerting to fans of various Ethernet-over-copper access technologies, which promise to extend the reach and expand the bandwidth of existing DSL technology to promote broadband access. In the case of Valo, at least, it appears that investors got cold feet.
Hatteras, an Ethernet access startup, was developing technology that could route Ethernet over a rage of access media, including copper and fiber. CEO Kevin Sheehan won't say how many folk were let go, nor would he disclose the current census. Sheehan says the company is still "healthy," well funded (amount undisclosed), and working through trials with major North American incumbents. But several sources described the layoff as a substantial cut.
Valo closed down this week and sent its 50 employees packing. CEO David Stehlin says he can't comment on what Petaluma, Calif.-based Valo is doing to sell its technology or find additional financing. "All I can say is that this team and this technology are definitely worth saving," he says.
Stehlin says Valo's current investors were supportive and the company was in the process of closing a funding round. However, some new investors that were to be a part of the round "squashed the deal" just moments before pen was put to paper, he says.
Valo wasn't yet in product trials, but the company had attracted notice because of its founder, George Hawley. Hawley, the chairman of Gluon Networks Inc., was also the founder of Diamond Lane Communications, which was acquired by Nokia Corp. (NYSE: NOK) in 1999. Prior to Diamond Lane, Hawley seemed to be involved in all sorts of broaband technology except perhaps inventing the Internet.
Stehlin says Valo's product was an access system that was about one-foot high and could serve about 400 copper pairs from a carrier's central office. The system was designed to use a carrier's existing copper lines in order to provide business customers an alternative to buying T1 lines (24 voice channels at 1.5 Mbit/s) or breaking the bank on T3 service (45 Mbit/s, or 28 T1s) for their local area networks.
"There's not a great platform deployed today," he says. "The platforms that are in the network today will run out of gas pretty soon. If you have a great platform and you have a great bonding protocol, you can get north of 9 Mbit/s downstream to every residence that has copper. You can do a lot of damage and bring in a lot of new revenue."
Hawley and company garnered some attention because of their work with standards groups to create a bonding protocol -- known as Inverse Multiplexing over DSL (IMD) -- that is more advanced than today's inverse multiplexing standards, such as inverse multiplexing over ATM (IMA), the physical layer technology that calls for ATM cells to be broken up before (and reconstructed after) traveling across T1 links.
"IMD builds upon the existing strengths of ATM standards, while providing efficient carriage of Ethernet, ATM, Frame Relay, and TDM traffic," states a technical document from the T1E1 subcommittee of the Alliance for Telecommunications Industry Solutions (ATIS). "It is inherently protocol independent, flexible and extendable."
In short, Valo and its competitors are trying to find ways of cramming more data down existing copper pairs and, when appropriate, bonding several pairs together to form bigger bandwidth pipes.
So what's going on here? Have the venture capitalists decided that copper is passé? Though the RBOCs have given quite a bit of attention to competitive access technologies, such as fiber-to-the-premises (FTTP), these vendors insist that there is a lot of interest in using what's already in the ground.
"FTTP is a very slow build," says Stehlin. "It takes a very long time and its very expensive. If you can reuse the copper you have, [carriers can] serve 80 percent of the homes in their reach right away."
Interestingly, Valo's competitors mourned the startup's plight, but they also had their guesses as to why it had such a hard time finding funding. "Bonding several copper pairs is not easy," says Bill Szeto, CEO of Ceterus Networks. "Some approaches, such as using DSL, have distance sensitivity," he says. This issue, and the fact that copper pairs don't have uniform transmission characteristics, can force product costs up, making sales tougher to close, he says.
The competitive set are also quick to point out that Hatteras and Valo may have been unique in their setbacks. "Actelis's progress continues, with September already a record sales month," writes Thomas Reynolds, senior VP of worldwide sales and service at Actelis Networks Inc., in an email to Light Reading. He claims a "growing sales funnel, repeat customers, and trials in PTT labs across the world."
Ceterus, too, says things are dandy. It has Global Crossing Holdings Ltd. and Progress Telecom as customers, but it admittedly survives "on a shoestring," Szeto says.
"Healthy competition is a good thing," Stehlin says with a sigh. "What's bad is the funding market."
— Mary Jander and Phil Harvey, Senior Editors, Light Reading