Sand Hill Road's Networking Guru Plugs Us In Vinod Khosla on Cisco, Redback, Juniper, and more.
By Adam Lashinsky

(FORTUNE Magazine) – No doubt Cisco Systems sits atop the new-economy hill. The question now: Is there a company out there capable of pushing it off? It happens that there's one man uniquely positioned to answer the question--Vinod Khosla, a Silicon Valley venture capitalist, a major Cisco shareholder, and a participant in just about every company that could give Cisco a run for its money--especially if they were to get together for the effort.

It's no surprise if you haven't heard of Khosla, who's less well known than his ubiquitous partner, John Doerr, the unofficial spokesman for the Internet economy and the face of Kleiner Perkins Caufield & Byers. But if Doerr is closely associated with Kleiner investments like Sun Microsystems, Netscape, and, Khosla will probably be remembered as the venture capitalist who helped create the next generation of the networking industry. Indeed, because of his intriguing involvement with numerous would-be Cisco killers, any combination of networking leaders joining forces to oppose Cisco would likely have to kiss Khosla's ring.

What Cisco offers--but individual companies do not--is one-stop shopping for high-speed telecom networking gear. In industry parlance, this is an "end-to-end solution," and it's arguably a requirement for any company that plans to be around for a very long time. But the companies Khosla has backed could--in the aggregate--offer something close. He founded and funded Cerent, the optical-networking company Cisco bought last year for more than $7 billion, hence his status as a Cisco shareholder. But he's also on the boards of router maker Juniper Networks and Redback Networks, which makes equipment for phone companies offering broadband DSL service. The two already compete with big telecommunications network operators, and together they could be a formidable foe of Cisco's. Throw in Cerent competitor ONI Systems, on whose board another Kleiner partner sits, and you'd have a bona fide Cisco look-alike, albeit one with gargantuan integration challenges. The pies in which Khosla finds his fingers don't stop there. He's also a director of Corvis, the supplier of optical-networking equipment to telephone carriers, which is expected to go public by early August.

And in case you think Khosla understands only the perspective of the gearmakers, he sits on the board of the combined Qwest and US West as well, perhaps the best example of a company building new telecom networks and a major purchaser of equipment from most of the above-mentioned manufacturers. With all this as backdrop, it only makes sense to try prodding Khosla into discussing the likelihood that "his" companies will join forces.

While dodging the specifics of consolidation, Khosla suggests why he's likely to be cool on the idea. "Any smart network operator is planning on having its most important equipment come from startups," he says, noting that Qwest in particular has bought equipment from Cerent, Juniper, and Redback. "From a network operator's perspective, you find that 'best of breed' is best."

Khosla's not talking about golden retrievers here. He's referring to the industry practice of buying each component from the top supplier. It's akin to selecting individual stereo components from different manufacturers rather than choosing an all-in-one system. And that is the fundamental debate affecting consolidation. "Best of breed" suppliers can beat Cisco product by product. Cisco's main advantages are large-volume deals and, as mentioned above, one-stop shopping.

A get-Cisco strategy "has its pros and cons," says Mike Volpi, chief strategy officer and top dealmaker for Cisco. "If they're successful in pulling off a transaction, it gives them a more complete, end-to-end product offering." But when Cisco's competitors try combining, they "usually screw up," he says, citing the merger of Synoptics and Wellfleet to make Bay Networks (which was later swallowed up by Nortel), Ascend's acquisition of Cascade Communications, and Lucent's subsequent acquisition of Ascend.

Khosla, who was born and educated in India and co-founded Sun Microsystems before joining Kleiner in 1986, started placing bets on equipment makers for new networks about three years ago. He has stuck with Internet technology instead of old telephone technology, and today he says he's focusing on the premise that broadband-access costs will continue to plummet. Khosla forecasts that in three years a high-speed corporate T-3 connection will cost $1,500, compared with upwards of $30,000 today. That's why he's looking at service companies that will sell revenue-enhancing services (like voice calls over data lines or network-hosted business software) and take advantage of falling access costs.

"Hardware is more polar than services," says Khosla, meaning there's no middle ground. A new networking product either works (and sells) or doesn't (and flops). "Services require more patience," he says. "But in the end you have more recurring value." Kleiner-backed offerings in this arena include application service provider Corio (which Khosla declined to discuss because at press time it was in registration to go public), Asera, Broadband Office, and OnFiber Communications, a recently formed venture that offers bandwidth time for corporate customers.

Khosla has a vested interest in expressing this opinion, of course, but he believes startups have a built-in advantage over established players. "It's not often you can displace a big player in an existing marketplace," he says, noting that Sun, Cisco, and Microsoft each created a new market rather than move into one in which another company was dominant. And Cisco, for one, has learned a few choice lessons that previous generations of giants never did. The biggest? If you can't beat 'em, eat 'em. The $480 billion company has proved itself to have a voracious appetite and is likely to gobble up any and every startup that shows promising technology in the field. Yes, such a strategy can lead to paying too much for too little, as some analysts have complained. But right now Cisco's plan seems to be working.

So does Khosla see any soft spots in the the equipment company's defenses? Can he point to any looming coalition of networking firms ready to slay the giant? "I don't even want to be quoted talking about it," says the normally loquacious venture capitalist. "People just read too much into it."

ADAM LASHINSKY is the Silicon Valley columnist for