Optical/IP Networks

Alcatel Boxes Clever

If it’s Tuesday, it must be another terabit router.

Alcatel SA http://www.alcatel.com (NYSE:ALA) is the latest in an interminable line of equipment vendors to dive in, late, to an already saturated market for super-routers (see Too Much Terabit? ).

Bad idea? Not necessarily. It turns out that Alcatel’s new product is just one element in a cunning plan to enable a new breed of bandwidth resale services over the core of the Internet – one that has the potential to boost Alcatel’s sales of both routers and optical crossconnects.

Alcatel’s router, called the 770 RCP (Routing Core Platform), will not be officially announced until September, but the vendor gave Light Reading a preview.

The basic tenet of Alcatel’s strategy, and the thing that sets it apart from many of its terabit competitors, is that it intends to the sell the 770 as part of an integrated solution set that also includes its optical crossconnects and an overarching suite of management and service provisioning software.

That integration is important, Alcatel says, because it allows it to move past the enabling of basic IP services to deliver virtual private services sold by “backbone wholesalers” over the Internet core.

This is a hot emerging market. The idea is that long-distance carriers with IP networks -- such as Qwest Communications Inc.http://www.qwest.com or Williams Communications Group http://www.williams.com -- will build core networks using optical crossconnects or switches, then surround them with a “perimeter fence” of routers. The routers will use the MPLS (multi-protocol label switching) traffic engineering protocol to divvy up capacity and allow the wholesalers to make money by selling it on, in smaller increments, to service providers like CLECs, or even very large enterprises.

Alcatel is far from being the only vendor shooting for this goal. MPLS nirvana is currently being sought, to some degree, by all of the networking juggernauts – including Cisco Systems Inc. http://www.cisco.com (Nasdaq:CSCO), Lucent http://www.lucent.com (NYSE:NT) and Nortel Networks http://www.nortel.com (NYSE:NT).

What makes Alcatel think it has a hope against the competition? “Enabling backbone resale services requires both strong routers and strong crossconnects, as well as the management to keep it all in sync,” says Dirk Van Den Berghen, vice president, data core, at Alcatel Bell NV’s carrier internetworking division. “Nortel and Lucent are strong in optical crossconnects, but weak in routing. Cisco is strong in routing, but doesn’t have a crossconnect. We intend to be the only vendor with all the pieces required to complete this service provisioning puzzle,” he adds.

It’s half-way there, Alcatel says. The vendor claims it already owns 37 percent of the US market for optical crossconnects (although at press time Light Reading was unable to find independent market research to corroborate that claim). And it’s formed a team to build a suite of management and control layer software, code-named Optip, to provide a one-stop point of control for both its routers and crossconnects.

The big question is whether Alcatel can really deliver on its promise of a high-performance, scalable, ultra-reliable, carrier class router.

It’s certainly talking the talk. Alcatel claims the 770 has been in development for a year and a half, and will be in customer trials by the end of the year. In the first iteration, the device will support up to 120 OC-48 ports -- or 120 Gbit/s aggregate capacity, it says. It also claims to have developed a “unique” switching fabric that allows the device to scale to “dozens of terabits of capacity.”

But even if that’s true, hardware is only half of the solution. Developing reliable and high-performance routing code is an even bigger challenge. (In a recent interview with Light Reading, Rafat Pirzada, co-founder of Cyras Systems Inc. http://www.cyras.com, said that one reason why his company had chosen not to add IP routing to its optical switch was because of the almost impossible task of finding people capable of developing carrier-class IP code. “There are only four individuals in the industry capable of that feat,” he said -- see Cyras: The Next Cerent? ).

Alcatel must also overcome another problem: abysmal marketing. It’s done such a poor job telling its story to date that some of its customers admit to buying its products in spite of its marketing message, rather than because of it (see Jack McMaster, CEO of KPNQwest). It’s also worth noting that news of the 770 emerged only after a group of senior Alcatel personnel organized a guerilla marketing action -- making an end run around the company’s official US PR channels to get the story to Light Reading.

Looking to the future, Alcatel’s biggest competition could come not from optical networking giants like Lucent or Nortel, but from two smaller companies that eventually could form one entity -- namely Sycamore Networks Inc http://www.sycamorenet.com, and Tenor Networks Inc. http://www.tenornetworks.com. Sycamore makes optical switches for core networks. Tenor makes an MPLS switch that, like Alcatel’s 770, is designed to allow long-distance carriers to resell bandwidth on an optical core network. Industry scuttlebutt has it that Sycamore is likely to acquire Tenor.

-- Stephen Saunders, US Editor, Light Reading http://www.lightreading.com
Be the first to post a comment regarding this story.
Sign In