But even though analysts say Atrica's first out of the gate with its particular set of features, the startup faces stiff competition -- and possibly a hard sell to incumbent carriers.
To analyze the situation, let's start at the top: Atrica has released a new version of its A-8800 platform that switches Ethernet traffic onto DWDM channels, based on a dynamic selection of best paths made by the switch's network management system.
Atrica claims its new gear switches traffic associated with up to 500,000 distinct MAC addresses, at 10 Gbit/s over 32 DWDM channels; it supports bidirectional Ethernet over lambdas in point-to-point or point-to-multipoint configurations at distances to 300 km; and it can be used in metro core or access networks to create VPNs or so-called transparent LAN services.
The vendor maintains that it's the combination of Ethernet switching and DWDM in one box that makes its switch unique. By putting the two functions together, carriers substantially expand their Ethernet capacity while saving space and transponder costs.
"I do think Atrica's first to offer this particular package of features in an Ethernet-based platform," says Michael Kennedy, managing partner at Network Strategy Partners LLC. "I haven't seen the whole thing -- ring integration, DWDM, SLAs, and quality of service -- supported all together this way yet."
Still, Kennedy points out, Atrica faces lots of competitors and it may not have as wide an appeal as it would like -- at least not initially.
Indeed, there's a wealth of products that purport to help carriers create Ethernet services -- many announced at Supercomm 2002 this week (see There's an Ether-Fest at Supercomm). Each has its own approach and set of advantages.
In general, these products fall into two main categories, which are useful to look at when analyzing what Atrica is trying to do:
Players in the Ethernet camp appeal to newer carriers that are just building out their networks, or to incumbents willing to try new techniques to branch out. These solutions are often cheaper and easier to manage than Sonet gear.
Sonet-based solutions appeal to carriers that are wedded to Sonet, know and trust their existing providers, and nurse the perception that purpose-built Ethernet solutions may not have the carrier-class features incumbents need.
Bottom line? Atrica's probably going to appeal most to greenfield carriers -- at least until it can prove itself to more established service providers. Unfortunately, those may not be the players that will be able to offer Ethernet services on the scale Atrica has in mind. Indeed, the telecom landscape is littered with the corpses of fallen startups that failed to garner a significant volume of customers (see Yipes Files Chapter 11: Is Ethernet a Sustainable Business Model?).
That said, Atrica's got the backing of a series of Tier 1 carriers (see Atrica Attracts $75M), including BellSouth Corp. (NYSE: BLS), Bezeq, and France Telecom SA. Allies like these virtually guarantee at least one win that could help boost the startup's credibility and presence.
— Mary Jander, Senior Editor, Light Reading
http://www.lightreading.comFor more information on Supercomm 2002, please visit: Supercomm Special
Atrica claims its new gear switches traffic associated with up to 500,000 distinct MAC addresses, at 10 Gbit/s over 32 DWDM channels...
Can someone help me understand where this would ever be required? I realize that many of the top markets may have millions of IP devices, hundreds of thousands of which may even want to VPN to a remote location, but surely the vast majority of businesses (small, medium or large) are going to VPN the networks together, not each and every computer.
Where is the market for this? If they (like Yipes!) are going to depend on the economy of scale, how are they ever going to get a medium market, let alone a mass market?