PacketLight's Pure Packet Play
”What, another one?”
That's likely to be the popular reaction to yet another new player entering what’s already considered (rightly) to be the most over-crowded market in networking.
However, PacketLight says its all-packet approach will set it apart from the hoi polloi. Simply, it claims to have developed a switch fabric that handles an array of applications (data, voice, video) and protocols (TDM, ATM, IP) – all via packets.
This approach supposedly delivers some big benefits: Allowing one type of interface card to be used to service ATM (asynchronous transfer mode), IP (Internet protocol), and TDM (time-division multiplexing) connections is cheaper than using different cards; and allowing carriers efficiently to "pack" those protocols into lambdas is another money-saving ploy. All-packet networks will also be easier to plan, provision, and maintain, the company says.
It’s a good pitch. The only problem is, PacketLight is not the only company making it. Ocular Networks Inc., for one, also claims to use one switch fabric and one flavor of line card for all traffic (see Ocular Announces First Product). And Ocular readily admits that it is not alone. “Most people do use different switch fabrics. But some don’t. And some other companies do offer multiprotocol cards,” says Doug Green, vice president of marketing at Ocular.
Money may be an issue for PacketLight. The company has received $29 million in funding and openly admits that it is unlikely to receive more in the short term. “It’s a tough time to get money right now. We have to wait until we have proof of concept before getting our next round,” says Yaki Luzon, vice president of marketing at PacketLight. That may not happen until Q1 next year, when its products are due to start shipping. PacketLight is currently burning about $1 million a month, Luzon says (which seems low, given that it has 85 employees).
PacketLight is backed by ADC Telecommunications Inc. (Nasdaq: ADCT), which owns 30 percent of the company. However, PacketLight execs deny that their company is an ADC "spin-in" (a startup that is set up and funded by a larger company with the express intention of being folded into the parent at a later time). In fact, they deny that an acquisition by anyone is in the cards. "We've all had our share of acquisitions. I don't see that happening any time soon," says Luzon.
In the end, the winners in the metro market may not be the companies with the best technology, but rather those that bring their products to market the fastest, and have the most funding. And there are several metro startups with a significant lead over PacketLight in that respect – including not only Ocular, but also Appian Communications (see Banks Pile Into Appian) and Coriolis Networks Inc. (see Coriolis Bursts Into SANs).
PacketLight is announcing two products at SuperComm: the SiteLight platform is installed at the edge of the network, on the customer premises or in a carrier’s POP (point of presence). It collects traffic from various services and protocols and multiplexes them onto lambdas on a metro fiber network. They’re then handed off to the second platform, CentraLight, which breaks out the different types of traffic and distributes them onto the appropriate core network (ATM, IP, Sonet, or so on).
For more information on Supercomm 2001, please visit the Light Reading Supercomm 2001 Preview Site.
— Stephen Saunders, Founding Editor, Light Reading http://www.lightreading.com