Optical Networking Space Rocks with $7.4 Billion in Cisco Acquisitions

By Erik Kreifeldt

Contents
Revolutionary SONET vendor
Critical link
Missing links
No blue light special
Another link
Cautious optimism

With the $7.4 billion acquisition of access add/drop multiplexer (ADM) start up Cerent Corp. (Petaluma, CA) and cross-connect start up Monterey Networks (Richardson, TX), Cisco Systems (San Jose, CA) sets the foundation to build a complete optical networking portfolio. The move also shows that Cisco, a staunch proponent of pure Internet protocol (IP) network architectures, needs traditional time division multiplexing and transport engineering competence to compete in the network equipment market.

"IP vendors have realized that next-generation networks are not just about routers and fiber," says Tim Weingarten, research analyst with BancBoston Robertson Stephens in New York. "Transport is an important piece of the puzzle, and next-generation transport will not be built by traditional vendors."

As rivals, customers, and analysts evaluate Cisco's strategy, they speculate how this and subsequent moves will effect Cisco's ability to compete with traditional telco suppliers and fend off or absorb challenges from innovative start-ups in the optical networking space.

Revolutionary SONET vendor
At $6.9 billion, based on the closing price of Cisco stock on August 25, Cerent becomes Cisco's largest of 40 acquisitions. In addition to buying Cerent for 100 million shares of stock, Cisco will shell out 7.3 million shares for wavelength cross-connect start up Monterey Networks, worth $500 million. Cisco previously had investments in both companies.

Cerent has posted less than $10 million in sales in its 2.5-year history. But the company sports 100 customers for its 454 product (see Data-Efficient SONET/SDH Platform Rolls Out at 240 Gbps), the only advanced technology to challenge the synchronous optical network (SONET) access multiplexer market since the early 1990s. The product rivals traditional gear with lower prices, denser port counts, and more efficient transport of data traffic.

Critical link
"The biggest thing that Cisco needed to compete effectively in the marketplace was transport products—an ADM," says Aman Kapoor, optical networking analyst at RHK (South San Francisco, CA). "The best one out there available for sale was Cerent."

Telco equipment titans Lucent Technologies (Murray Hill, NJ), Fujitsu Network Communications (Richardson, TX), and Nortel Networks (Brampton, ON) dominate the SONET industry. Industry analysts at RHK forecast that the SONET market will grow more than 60% in 1999 to $7.3 billion.

"I think it's a good move for Cisco to enter the optical networking business," says Wayne Price, chief technology officer for Williams Network (Tulsa, OK). "Cerent was very well positioned to erode the ADM market from incumbents like Lucent, Nortel, and Fujitsu," he says. Price's company backs up his sentiment with a $25 million purchase order for Cerent gear announced in June 1999.

But $6.9 billion for a start up with less than $10 million in revenue for its first six months? RHK's Kapoor offers a comparison between Cerent, which recently filed for an initial public offering, and newly-public network plays Juniper Networks (Mountain View, CA) and Redback (Sunnyvale, CA).

Filling in another piece of the network puzzle for Cisco is Monterey. Driven by the proliferation of dense wavelength division multiplexing (DWDM) systems in backbone networks, the company is working on a cross-connect that switches and routes optical signals. It expects to ship a system with 256x256 OC-48 ports early next year. In addition to traditional telco rivals, start-up Tellium (Oceanport, NJ), in which Cisco also has an equity stake, is working on a similar product.

"Monterey has the potential to be as exciting as Cerent," says BancBoston's Weingarten. Monterey sports a formidable product strategy and brain trust, but unlike Cerent, Monterey has yet to announce any customers or ship commercial product. Tellium was first-mover in the optical cross-connect space, and boasts a $250 million supply contract with Extant Communications (Aurora, CO) for its first and second generation gear.

The purchase of Monterey also marks an about-face in Cisco's technology vision. The company previously advanced the philosophy that IP networks perform best when routers have full access to transmission capacity, and that cross-connects and SONET ADMs add unnecessary complexity to the network.

Cisco has publicized its efforts to link IP routers directly to DWDM equipment by developing SONET interfaces for the routers, sparing carriers the need to use expensive SONET ADMs or cross-connects in between. Most carriers, however, balk at building national-scale mesh networks of IP routers linked with fiber and DWDM channels without the kind of traffic engineering that cross-connects and ADMs provide.

With Monterey equipment, Cisco gains competence sought by carriers. Cisco positions the Monterey and Cerent technology as enabling service providers to migrate from traditional circuit-based networks to cell and packet-based networks.

Missing links
As if $7.4 billion worth of optical network competency were not enough to ponder, industry experts point out what Cisco did not get for its money, and speculate about what the company might do to get it. The missing link is DWDM—both long and short versions.

The products from Cerent and Monterey represent two of four fundamental components of optical networks, including long-haul DWDM, large-scale optical cross-connects, metro transport, and access. Cisco gained access and cross-connect technology with Cerent and Monterey, respectively.

"Cisco hasn't solved the transport puzzle by buying these two companies," says BancBoston's Weingarten. "The most glaring hole in Cisco's portfolio is metro transport." In line with the way Cerent revolutionized local access equipment, bringing in a fresh approach to metro transport would round out Cisco's optical product line between the Cerent box and the Monterey device.

Cisco already has investments in San Jose neighbor Optical Networks, a metro DWDM start-up. "Optical Networks is really defining the metro transport space," Weingarten observes, noting that initial metro DWDM efforts from Ciena Corp. (Linthicum, MD), Lucent, and Nortel (via Cambrian) are not catching on—and are subsequently leaving open an underestimated opportunity.

No blue light special
"The last thing we want to do is to get purchased. We do not want to be another division of Cisco Systems," says Optical Networks CEO Hugh Martin. "The reason I'm here is for a chance to go build another Cisco," he proclaims, noting that with revolutionary technology, management resolve, and capital resources, Cerent, too had the gumption to become the next Cisco. "In the end, they blinked," Martin asserts.

While distancing his company from an acquisition target, Martin acknowledges that the Optical Networks product platform fits snugly between that of Cerent and Monterey. Optical Networks has established interoperability between its product and that of Cerent, and executives from both companies engage in joint sales calls, he reports. Optical Networks also has interoperability work underway with Monterey-like bandwidth management equipment, he adds.

Speaking of interoperability, Cerent and Monterey have also worked together (see Monterey Launches Wavelength Router, Demonstrates Interoperability with Cerent). The collaboration gives Cisco a head start on design team collaborations, but experts say carriers will prefer to deploy metro transport equipment in between Cerent- and Monterey-like platforms.

Missing another link
Another big link in the optical network chain lacking in Cisco's new arsenal is a source of long-haul DWDM to link hubs of its Monterey cross connects. Cisco already works with long-haul DWDM stalwarts Ciena, Nortel, and Pirelli Cables & Systems North America (Columbia, SC). In keeping with a next-generation approach, start ups such as Corvis (Columbia, MD) and Qtera (Boca Raton, FL) represent this long-haul optical space. Cisco has an investment in Corvis.

Contrary to Martin's bold reaction to the news, Cisco's primary rival for multi-gigabit IP routing, Juniper, professes no interest in matching Cisco's move. "We're focusing on offering complete solutions for ISPs [Internet service providers]," says Juniper spokesperson David Abramson. Juniper already offers the highest capacity commercial router, along with the requisite software prowess, service capability, and port density to outfit Internet backbones, he says, adding that the company has no plans to match Cisco's efforts for end-to-end public networking by acquiring companies.

Cautious optimism
The prospect of consolidation in the optical networking industry leaving market dominance in the hands of a few large players does not bother Williams' Price. "I believe you'll always end up with two or three large players," he says, adding that the number generally provides enough competition among vendors to ensure favorable price/performance ratios.

"Williams is one of the most aggressive carriers in dealing with start-up companies," Price asserts. One problem the carrier encounters is the new companies' ability to scale up manufacturing capacity. When startups start to bring on multiple customers, they often can't keep up with demand, Price says. "Cerent probably won't run into that problem with Cisco."

While generally happy about the deal, Price remains cautiously optimistic about its execution. Product development progress often slows down for six to 12 months while merger activity takes place, he notes. Price would hate to see Cerent, which is executing an aggressive and attractive product development plan, get derailed by the acquisition process. But he takes solace in the notion that Cisco did not likely shell out billions to acquire a company, only to let it falter.