) the Microsoft of networking, and then he helped Juniper Networks (JNPR
) become Cisco's top rival in the market for the world's fastest routers.
Later, Li co-founded Procket Networks, which raised more than $300 million in funding during the Net boom, in large part due to Li's involvement. But Procket never took off, and Li left shortly before it was bought by Cisco in 2004. Shortly thereafter, he rejoined his former Procket mates by returning to Cisco -- bringing him full circle.
Now, BusinessWeek has learned that Li, 43, is heading back to startup land. He departed Cisco on Sept. 16 to join Portola, a tiny outfit whose CEO is Vito Palermo, Procket's former chief operating officer.
SOLVING THE WEB. For the first time in his career, Li is not focused on developing a super-fast router, the machines that zip all manner of Internet traffic around the world. Tools such as routers, which form the basic plumbing of the Internet, are already sufficient, Li says. So now he is focusing on a new problem: developing software to help consumers get whatever they want off the Net, and have it delivered to whichever devices they desire.
"What I'm really trying to accomplish, in the big picture, is to find more good uses for the Internet," Li says. "There's an entire communications infrastructure out there, and we've just begun to explore how to use it. The Internet is too complicated. We should be able to get to the point that a 5-year-old can run it, as effectively as a 45-year-old geek."
If Li succeeds, consumers will automatically get their digital entertainment -- be it a song, TV show, or movie -- delivered with no fuss to any or all of their gizmos, from cell phone to PC to living-room TV. Think: no more configuring networks, messing with incompatible data formats, or synching up devices.
BOLD GOALS. Portola's software would manage all of it. For instance, a purchased movie would be transferred in good-enough resolution that it could quickly be downloaded to a wireless multimedia phone, or in full high-definition grandeur over a broadband connection to a big-screen plasma TV. "All of these things should happen auto-magically," says Li.
Palermo says Portola will first focus on developing this "Internet Entertainment Delivery System" to handle movies, followed by music videos and then computer games.
Of course, this is big talk for a six-person outfit that has only a rudimentary working prototype. Moreover, technically speaking, it's an audacious goal. Portola's software would have to handle content coming off varied networks (say, fiber, copper, or wireless), in multiple formats (say, to work with Microsoft's Windows Media Player and RealNetworks' RealPlayer), and with the firmware and drivers needed to jibe with all kinds of devices. "It's the integration of everything," says Li.
"BIG RUBIK'S CUBE." Audacious or no, Portola's aim is likely to resonate with a wide swath of potential customers. If it can deliver, Portola could help content providers such as digital-music services and game companies to cut networking costs, while increasing the speed and reliability of delivery.
"This can become a very significant way to deliver content," says Frank Dzubeck, founder of consultancy Communications Network Architects. One game-publishing executive, who requested anonymity, notes that his carrier partners demand 30% to 50% of the revenue for all games sold via their websites. Portola would charge far less, says Palermo.
More intriguing is the potential for Portola's approach to allow content owners, such as movie studios and record labels, to sell directly to consumers. That's not the way it works now, when they sell their wares via distributors, such as phone companies, portals such as Yahoo.com, or Internet e-business outfits such as Napster (NAPS
) and MovieLink.com. Why? Most content owners lack the technical knowhow.
"The reliable delivery of content is a big Rubik's Cube for content publishers," says Keith M. McCurdy, CEO of Kayak Interactive, which makes games that can be used on mobile devices. A former executive at games giant Electronic Arts (ERTS
), he points out that most content providers are experts at licensing, marketing, and managing their properties -- not at ensuring it can be delivered to every conceivable device anywhere on the planet. "If someone could come along and make that process reliable and effortless, that would be significant," McCurdy says.
MY NET. Portola reckons it has a plan to do just that. It would buy the necessary bandwidth from carriers to distribute customers' content, through one of nine data centers it plans to co-locate in Internet hubs by 2007. Because of the company's special software and economies of scale, Palermo thinks Portola should be able to provide a 30%-plus cost savings, vs. the in-house approach. Then, when a consumer purchases a song or movie, Portola would get a cut.
What would it mean for consumers? If this direct approach pans out, they would be able to search for precisely what they want, without having to opt for prepackaged offerings. If the only program a viewer watches on TV is The Sopranos, he or she could order it directly from HBO, without having to buy that pricey premium bundle from the cable company.
Phone, cable, and Internet companies have spent recent years gearing up for so-called bundle wars, in which they would all compete with broad offerings including phone, TV, and data services. But many tech analysts believe these carriers would be better off sticking to what they know best: providing basic connectivity. Rather than pouring marketing dollars to hawk these bundles, they should cut costs to make a good living by selling basic pipes. One reason: today's Net-savvy youth are far less brand-loyal, and are comfortable Googling to find the precise content they want.
REALITY OR SPIN? Focusing only on getting content into a carrier's bundle is "very old-school thinking," says Li, who points to the success of Apple Computer's (AAPL
) iTunes Music Store. The online store, which is actually a client application downloaded to users' PCs, gives Apple a direct relationship with its customers, rather than having them buy music from SBC Communications (SBC
), Comcast (CMCSA
), or Yahoo.com (YHOO
). Says Palermo: "Today, consumers are relying on carriers to make choices about what they can see. I'm looking at a totally open Internet-based model, where the consumer decides what he or she wants to access."
That's not to say that carriers will get out of the bundling business -- or that content providers would avoid them even if they could. Most content owners will still seek partnerships, because carriers can give them instant access to millions of consumers who might not otherwise know how to reach them.
And Portola is not going to strike any fear into anyone anytime soon. It hasn't even raised venture funding as yet. And there are those who think Li's fame is overdone, the result of PR campaigns designed to burnish his rep. "Tony is one of the few people I know in the router industry that actually takes credit for doing something. He decided a while ago to stand out," says Dzubeck.
A STEP AHEAD. Li's former boss, Juniper CEO Scott Kriens, says Li is no self-promoter, though he's not afraid to voice his point of view. "He's a really smart guy, and he's really passionate about his beliefs," Kriens says. "Because he has a lot of opinions, he gets a lot of opinions back -- including about him."
Something else Li believes: "There is about $100 billion of content that's created each year, and only 1% of it is distributed digitally," he says. "We're banking on the fact that this is a wide-open field." And it's that kind of thinking that may take him far. Notes Kriens: "He's always trying to think about where the world is going, and what customers need -- about making technology that's relevant to the real world."
Burrows is Computer editor in BusinessWeek's Silicon Valley bureau