Celox Networks, an IP service switch startup, closed its doors for good this week. Yesterday, the company took down its Website and posted a message for all to see:

The company burned through a total of $155 million of venture funding in roughly four years to create its SCx 192, an IP switch with an 80-Gbit/s backplane that the startup claimed could support OC192 packet processing and six million simultaneous users (see Celox Bags $80M Funding Round).

For a while, Alcatel SA (NYSE: ALA; Paris: CGEP:PA) was rumored to be interested in the SCx 192, but in the end nothing materialized. And although the switch made it to the AT&T Corp. (NYSE: T) labs, trials never generated any meaningful revenue. Celox also hoped to land a big contract with the U.S. federal government to build out a secure IP network, say sources, but the bidding process was extremely slow. With no other real prospects in the pipeline for the next six to 12 months, the venture capitalists backing the company were unable to continue funding.

Celox wound up laying off approximately 125 people in its Southborough, Mass., facility. It closed its development site in St. Louis this summer, laying off 90 people (see Meet Me in Southborough, MA?). “Celox is a case of simply dying on the vine, within eyeshot of a rising river,” says Scott Clavenna, president of PointEast Research LLC and director of research at Light Reading.

But some analysts say that Celox’s market was simply doomed. They say the need for IP service switches is waning, as much of the functionality is becoming incorporated into existing edge routing platforms. Cisco Systems Inc. (Nasdaq: CSCO) and Juniper Networks Inc. (Nasdaq: JNPR), for instance, have released various versions of software to create IP VPNs, aggregate traffic, and enable other IP services (see Edge Routing Gets Service Friendly ).

“There aren’t too many IP networks out there that don’t have Cisco routers in them,” says Curtis Price, an analyst with Stratecast Partners. “It’s much easier to upgrade what you already have than to deploy something completely new. Carriers are looking to cut operational costs and squeeze as much as they can out of what they have.”

Other startups targeting the IP service market have been struggling, as well. Corona Networks Inc., which makes an IP edge aggregation device, also recently laid off a large number of employees and is rumored to be on its last legs (see Headcount: Shopping, Lifting, Moving On). Ellacoya Networks Inc., which first marketed itself as a broadband aggregation and IP services company, completely changed its strategy and is now targeting the cable market with a smaller box focused on bandwidth policing (see Startup Renaissance in Cable Market).

Larger, public companies are also struggling in this market. Lucent Technologies Inc. (NYSE: LU) recently cancelled its SpringTide product (see Lucent Silences SpringTide). CoSine Communications Inc. (Nasdaq: COSN), which specializes in offering network-based VPN and firewall services, only reported $5 million in revenue for the third quarter of 2002 (see CoSine's Quest for Cash). Nortel Networks Corp. (NYSE/Toronto: NT) is still selling its Shasta product, though analysts speculate it isn't selling a lot of them.

Bert Whyte, director and CEO of Network Equipment Technologies Inc. (net.com) (NYSE: NWK), says carriers have been interested in service creation, but that most of the Scream boxes his company sells are for ATM and Broadband Remote Access Server (B-RAS) functionality.

“The reality is that no one is buying anything,” he says. “They are very interested in service creation, but I don't think they'll be deploying it until 2004.”

— Marguerite Reardon, Senior Editor, Light Reading