A couple of weeks ago, Cisco Systems Inc. (Nasdaq: CSCO) invited a group of Light Reading editors to its campus for a series of technology briefings and demonstrations. They even – mirabile dictu! – let us in their house and fed us.
(Heck, Cisco was so nice, they even took one of our editors to the hospital – due to neither the food nor fisticuffs! Senior Editor Phil Harvey threw out his back while sharpening his pen; but he's back in Texas and on the road to full recovery.)
A trip inside the Utopian world of "Cisco-land," as neighboring companies in Silicon Valley call it, can be unsettling. The Cisco campus sprawls over several parsecs, covered with dozens of identical, blocky, sand-colored buildings. It is especially bizarre considering that, after passing block after block of empty office buildings elsewhere in Silicon Valley, you discover that Cisco still has cars in the parking lot (see Real Estate Nightmare).
The real question, though, when contemplating the Cisco sprawl, is: How can this kind of growth be maintained?
To do so requires the ability to seize new opportunities. And in a series of meetings with several Cisco product groups and executives, it gradually emerged that one of the opportunities that Cisco is preparing for is multiservice technology.
Specifically: multiservice technology that can be used to construct metropolitan area networks. (Wireless networking and cable networks are also significant growth opportunities, the company thinks, but let's leave them for another time.)
In short, in service-provider networks, Cisco believes there is still demand for a flexible, multiservice switching and routing architecture that has yet to be built. "Carriers want a converged network, and nobody has it," says Mike Volpi, Cisco's senior vice president of its routing technology group (for more from Mr. Volpi, see our interview with him.)
The big, fat target according to Volpi, is something he calls "multiservice edge," which includes something that does "ATM, frame relay, traditional lease line aggregation, broadband aggregation, potentially metro Ethernet – all in one box."
Yes, you've seen this movie before. It's called Oh My God Box! (see God is Dead).
What does Volpi name as the closet product Cisco has? Why, a router, of course. Currently, Cisco would deliver a GSR, 10000, or 7600 router, to customers that require these features. While Volpi did not say so, it was clear to the Lightoids in the room that we can expect something more from Cisco in this multiservice space before too long.
To this end, Cisco is trying to "carrier-fy" its routing platforms. Most recently it designed new resilience features into its 7500, 10000, and 120000 series routers, which use use Nonstop Forwarding (NSF), Stateful Switchover (SSO), and Multiprotocol Label Switching (MPLS) Fast Reroute technology (see Cisco Intros Globally Resilient IP). These enable a router to run redundant processors and switch over from one processor to another with zero packet loss, according to the company.
But this isn't just a routing game for Cisco. Oh no. It is also attacking the multiservice issue from the transport side. Here, Cisco has a nice entrée into carrier metro networks with the immensely successful ONS 15454 (Cerent) platform. There are some 30,000 of these boxes deployed, according to Cisco. It plans on extending the reach of this platform with its new ONS 15600 metro optical switch (see Cisco's Fine Young Cannibal). This switch will be taking aim at Ciena Corp. (Nasdaq: CIEN), in the grooming switch market, and possibly Tellabs Inc. (Nasdaq: TLAB; Frankfurt: BTLA), in the crossconnect marketplace. If Cisco pulls off the ONS 15600 (which ain't done yet), it could further establish its presence in the metro POP.
We asked Rajiv Ramaswami, VP and CTO for the optical networking group at Cisco, how his employer planned to enhance its optical products. "Multiservice capabilities," was the reply. Ramaswami forecasts a wave of "...data interfaces added to scaleable Sonet/SDH/WDM platforms, supporting DS1, DS3, and SAN protocols." Cisco is also beefing up on technologies such as Resilient Packet Ring and MPLS, which can bridge the gap between data networks and the carrier transport side. RPR makes data look more Sonet-like, and it can be plugged into routers. This helps package data for carriers, which don't generally like things that don't look and smell like Sonet. MPLS? Well, it's a more router-friendly version of ATM.
Still, the transformation of large Tier 1 carriers networks into packet-switched cores isn't likely to be done for many, many years. In the meantime, they need to upgrade their ATM gear and build out the new multiservice edge.
Okay, so Cisco is trying to make its routers beefier – and they have a neat next-gen Sonet transport box with data interfaces. But where's the God box?
Here's the problem. Cisco is not a leader in either ATM technology or in the more general, carrier-class "multiservice switching" market. In fact, according to the July 2002 issue of Light Reading's paid subscription service, Optical Oracle – "Retro Chic: Multiservice Switching" – Cisco was running either third of fourth in the market, depending on how you count. It lags behind leaders Nortel Networks Corp. (NYSE/Toronto: NT) and Lucent Technologies Inc. (NYSE: LU) in the multiservice and ATM switching markets; it vies with Alcatel SA (NYSE: ALA; Paris: CGEP:PA) for the third spot, in the view of several market research firms.
The gist is that Cisco is still a packet-switching company, not a mulitservice-switching or transport company. The large RBOC carriers are moving to packet-switched networks, but very, very slowly.
It's not likely that Cisco will be able to get where it wants to go without substantial acquisitions. After all, that's how it got where it is now. Cisco isn't doing much shopping at the moment; but watch this space, as it is certainly thinking about it (see Cisco, Sycamore Circling Lucent's ATM) .
Cisco is biding its time, looking to make the right move for the right price. Where will it look? First, it will prey on the carcasses of Nortel and Lucent, scavaging for leftovers. Then, it will scope out the startup world and buy one or two of the hottest products at bargain basement prices.
Who'll it buy? Well, Volpi's already told us he doesn't like pure-play MPLS vendor Vivace Networks; and he's never heard of Appian Communications Inc. But it's likely to be one that has a strength in the ATM or broadband crossconnect market. Say, possibly, a WaveSmith Networks Inc.? Maybe a Polaris Networks? Tellabs?
It's important to bear in mind that Cisco has plenty of time. Unlike its major competitors in the carrrier market, Nortel and Lucent, its cash stash is not melting. Indeed, an analysis of Cisco's financial strength and product positioning led Optical Oracle to name Cisco No. 1 in our September ranking of "Top Recovery Candidates" (see Cisco, QLogic Top Optical Oracle Charts).
So, did Cisco executives brainwash us on our trip to San Jose? Not exactly. Cisco has work to do, and the competition from Juniper Networks Inc. (Nasdaq: JNPR) in the router market is nothing to clear one's sinuses at. But as a leader in the packet routing and enterprise switching world, Cisco is well positioned to make headway in carrier networks. Its doing that step-by-step, in a patient and methodical way.
Cisco's recent moves reflect the fact that the unfolding battle on the service-provider edge is the most important technology contretemps yet. IP and data are creeping deeper and deeper into the network. Carriers, believe it or not, are moving to packet-driven networks. Whoever wins the war for the multiservice edge will likely be the new carrier king.
— R. Scott Raynovich, US Editor, Light Reading
Editor's Note: Light Reading is not affiliated with Oracle Corporation.