Polaris Lifts Off
The company is aiming at the multiservice market, though it will do so in a series of steps. It wants to compete with entrenched crossconnect products such as the Tellabs Inc. (Nasdaq: TLAB; Frankfurt: BTLA) Titan 5500 system and Alcatel SA's (NYSE: ALA; Paris: CGEP:PA) 1631 LMC system. The first release of OMX supports Sonet services, such as DS1 and DS3 connections, with the ability to migrate to full multiservice capabilities, including support for a variety of Asynchronous Transfer Mode (ATM) and Internet Protocol (IP) services.
The OMX aims to one-up the old crossconnects, doing their jobs in a smaller space while adding the functionality of several Sonet add/drop multiplexers (ADMs). Its capacity starts at 240 Gbit/s and scales to 2 Tbit/s, and it handles circuits in VT1.5 (1.7 Mbit/s) increments versus core switches that only handle STS1 (51.84 Mbit/s) chunks. The product's interfaces will include DS3, OC3 to OC192, and Gigabit Ethernet, the company says.
"[When] positioned against the Tellabs and Alcatel DCS products, the Polaris OMX has the added benefit of being an ADM, so it will enable a significant reduction in the number of ADMs needed in a carrier's central office," says Maria Zeppetella, VP of network infrastructure at Probe Research Inc.
Indeed, cramming more functions into one box may, like corduroy, yet come back in fashion. With the amount of carrier consolidation happening these days, large carriers might be more prone to buying a dense metro box that can replace several standalone products, according to Scott Clavenna, president of PointEast Research LLC.
"Polaris will have to demonstrate radical cost and performance over competitors," Clavenna says. That'll be tough, but not impossible, given that many carriers need to scale their digital crossconnects because of increasing private line traffic, he says.
Polaris is indeed flogging its small size and density versus its competitors. In press briefings, the company claims its product would only take up one half of a telco rack at a cost of $1.6 million to provide 1,024 fully redundant STS1 ports. Polaris claims the Titan 5500 would require 27 racks of gear at a list price of $6.7 million to match that configuration. It says Alcatel's solution would require 29 racks of gear at $6.3 million.
Tellabs dismissed Polaris's comparison, saying its numbers were off "by an order of magnitude." "There is no configuration that we have ever sold that would be all OC12s -- that doesn't even make sense," says Greg Nulty, VP of product line management at Tellabs. Besides, with more than 4,000 systems installed, Nulty says "the likelihood that someone would replace a 5500 with something new is remote."
Alcatel confirms that its solution would require 29 racks of gear to match the configuration presented, but wouldn't confirm a price. "Startups such as Polaris are obviously compelled to make extravagant claims about their new products to get attention in the market," adds Jamie Horton, an Alcatel spokeswoman.
Of course, winning on paper will only get you so far. "They also have to demonstrate performance of the switch and financial viability of the company," Clavenna says. "[Polaris] may need a partner or original equipment manufacturer to get into big [carrier] accounts."
By putting Tellabs and Alcatel in its crosshairs, Polaris has actually narrowed its competitive field. When the company first announced its product architecture, it looked as though it was going up against everything from Sonet add/drop multiplexers and MSPPs or "God boxes" to the large core switches at the edge of long-haul networks (see Polaris Builds a God Box).
Startup Ocular Networks, which Tellabs bought last year, took on the Titan 5500 at one time (see Ocular Takes On Tellabs). However, once acquired, the Ocular box became complementary, and it is now marketed as a Titan 5500 Mini-Me for Tier 2 and Tier 3 carriers. Interestingly, Polaris says it is going head-to-head with the Ocular box (now the Tellabs 6400) in at least one carrier account, proving that despite its own marketing messages, Polaris will try to win business on either side of the metro core.
Polaris was founded in June 2000 and has raised $77 million to date without a down round, according to Cheryl Gray, Polaris's senior marketing manager. The company employs about 99 people, most of whom are engineers.
Several of its directors, including Ray Kao, the CEO and chief technology officer, were previously associated with Stratacom, the ATM switch company that Cisco Systems Inc. (Nasdaq: CSCO) bought in 1996. In contrast, a couple of Polaris's operating staff hail from Tachion Networks, a brash MSPP vendor that added too many functions in one box, failed to land a large customer, and went out of business (see Tachion Trolls for Cash).
Polaris has trials underway with one RBOC, one IXC, and two next-generation carriers, according to Sab Gosal, Polaris's director of product marketing. Three of the four carriers are using the box as a wideband transport platform, and one may be looking at it as a long-haul switch, he says.
Will carriers put money where Polaris's mouth is? Maybe, but not for a few months. If all goes well in its trials, Polaris could see initial product revenues by the second quarter of next year, Gosal says.
— Phil Harvey, Senior Editor, Light Reading