Offers its own approach to flow routing with its new core router

June 1, 2004

5 Min Read
Axiowave Queues in the Core

Long quiet, and almost forgotten, Axiowave Networks Inc. has suddenly popped out of the woodwork with, of all things, a core router.

Axiowave today is introducing the XCR128, which it calls a “service convergence router.” The company claims it will trump existing high-end routers, of which there are plenty, by increasing utilization rates of the router ports as high as 90 percent, using quality-of-service (QOS) engineering similar to what is employed in ATM switches.

The company is out to make IP routing more like the ATM and Frame Relay networking market, where service providers are able to charge a premium for guaranteed service-level agreements (SLAs).

Axiowave was founded in 2000 by CEO Mukesh Chatter along with Chairman Ray Stata, who is also the chairman of Analog Devices Inc. (NYSE: ADI). Chatter 'n' Stata also teamed up to found Nexabit, sold in 1999 to Lucent Technologies Inc. (NYSE: LU) for the now-legendary sum of $900 million.

Axiowave has raised more than $120 million from investors including Argonaut Private Equity Management LLC, Madison Dearborn Partners, and Soros Private Equity. Chatter and Stata themselves are also investors (see Chatter's New Box).

In many ways, Axiowave looks to be the “next-generation” Nexabit. Chatter contends that existing IP routers are only built for best-effort IP traffic, and that routers need an entirely new architecture. In the XCR128, he says, a 10-Gbit/s OC192 port can carry 90 percent premium SLA-based traffic, with the rest of the port dedicated to best-effort IP traffic. The entire router will scale to 1.28 terabits of capacity. It will support existing routing standards such as BGP, MPLS, and OSPF, but will also introduce its own software modifications for services such as "ATM-grade" IP-VPNs.

The router's main switching shelf is 35 inches tall by 21 3/8 inches wide by 23 1/4 inches deep, and it weighs 356 pounds when fully configured, Axiowave says.

”Incumbent router architectures are conceived to support best-effort traffic,” says Chatter. “We’ve designed a box to emulate the ATM/Frame Relay business model for IP.”

Axiowave's biggest challenge will be proving to the world it has a reason to exist. The core router market is packed with competition, capital spending in the market is depressed, and startups such as Caspian Networks Inc. and Procket Networks Inc. have already touted higher-performance architectures.

One analyst points out that both Caspian and Procket were focused on refining the performance of routers, and that they've still struggled to find new customers .

"I don’t think the latest crop of core routers missed the idea that you need to optimize around performance rather than scale,” says Scott Clavenna, chief analyst with Heavy Reading, Light Reading Inc.’s market research division. “They [Axiowave] are saying this is the first core router optimized for performance rather than scale… but Caspian and Procket aren’t lacking in IP QOS. What differs is how they each achieve it and how stable the router is in different networks and at different loads.”

Chatter says Axiowave will be unique when it comes to performance and throughput. Axiowave claims a major service provider has put the product through a battery of tests. The tests showed that high-priority traffic carried in a fully loaded switch fabric, with 90 percent utilization, results in a worst-case latency of 69 microseconds. He says today's current IP/MPLS routers can only dedicate 7 percent to 8 percent of any one port to premium services. Axiowave is not naming that service provider, but it says that it has a contract with a small Ohio-based service provider, PowerNet Global Communications Inc.

Proof of concept may have to come in the form of bigger customer orders, which may be tough, given that Axiowave is focusing on Tier 2 and Tier 3 carriers. Besides that, the core router market is filled with folks trying to compete with dominant leaders Cisco Systems Inc. (Nasdaq: CSCO) and Juniper Networks Inc. (Nasdaq: JNPR). These include startups like Caspian, Procket, Chiaro Networks Inc., and Hyperchip Inc. Then there's Avici Systems Inc. (Nasdaq: AVCI; Frankfurt: BVC7), a public company that's occupying third place, by most market-share measures, in the core market. Even incumbents such as Alcatel SA (NYSE: ALA; Paris: CGEP:PA), Nortel Networks Corp. (NYSE/Toronto: NT), and Nokia Corp. (NYSE: NOK), have dabbled in core routing.

Another big question is whether Chatter and Axiowave can overcome the baggage of Nexabit's past. Nexabit was intended to be Lucent’s next-generation core router, but that project turned sour after much internal melodrama and finger-pointing. Eventually, Lucent shut the whole project down (see Whatever Happened to X?, Lucent Loses Two Big Names , Lucent Quiets Terabit Router Rumors, Lucent Faces "Exodus of Nexabit Staff", and Lucent Cleans Up Core Routing ) .

Axiowave deserves credit for making it this far. The company started by building an optical crossconnect -- a box that would have been doomed from the start, given the plight of the optical switching market. Chatter says the company changed gears about two years ago, realizing the optical crossconnect market was not nearly as bright as the routing market. It went through its own cycles of refinancings, layoffs, and product changes (see Axiowave Closes on $45M and Stealthy Startup Leads With Layoffs).

”There are two angles to approach the Tier 2 and 3 markets: either smaller telcos or MSOs getting into triple play over an IP infrastructure, or data services specialists that are looking to offer IP VPNs with SLAs that match ATM and Frame Relay.” says Clavenna. “The thing to remember is that Cisco and Juniper are pounding away at those customers as well.”

— R. Scott Raynovich, US Editor, Light Reading

Subscribe and receive the latest news from the industry.
Join 62,000+ members. Yes it's completely free.

You May Also Like