When GIANTS Stumble

One direction vendors have scarcely considered yet — though many acknowledge as their eventual destination — is the migration toward software rather than hardware as true differentiators.

“Value is fleeing the network layer and moving up,” Nolle said. “So far, no one has really stepped up to be a strategic partner [to carriers] above the hardware layer. I'm looking to see whether any of these guys recognize the opportunity associated with the next layer up: the software and operations management layer. These big players have to look at extending beyond just the last service interface. In the main, they haven't done that.”

One of the few that has made the plunge, core router vendor Avici Networks, was perhaps in the best position to do so. With one customer left and sinking revenue, the status quo was riskier than creating a new business based on software control planes named Soapstone Networks.

As the migration of suppliers from hardware to software both corresponds with and accelerates the gradual commoditization of hardware, established vendors are likely to resist the trend with all their might to protect their legacy businesses. That would seem to imply a fertile environment for start-ups, whose role is best served when they're disrupting current architectures. But the venture capital community has cooled to the sector of late, opting more often to double down on existing bets than to make new ones — thus, the news last month of another $24 million in funding for optical transport vendor Tejas Networks, another $19 million for Mintera and, this summer, a return to the well for Calix.

According to PricewaterhouseCoopers, venture capital investment in telecom and networking gear decreased nearly 6% in the first nine months of this year to $8.3 billion. The number of deals in that period also dropped 6%, averaging less than 400 per quarter. The biggest venture capital deal of the quarter, a $52 million round by 9-year-old long-haul optics firm Xtera Communications, was a 10th round.

“The start-ups that will make it will have value-add beyond the current buildout cycles, looking not at the next three years but five to seven years out,” Hunt said. “U-verse is not going to change. It would be hard for someone to jump in there and disrupt it.”

Still, there's a disconnect between how radically services are transforming and how slowly and slightly network technologies are changing, Nolle said, which is creating another great opportunity for adventurous vendors. That opportunity means reversing the traditional bottom-up approach in which vendors, starting at the optical layer, declare what the network can do, and carriers figure out what services those abilities might allow.

“Another way to build the network is to look at what the [service delivery framework] can control and expose it: an operations-driven network rather than a bit-driven network,” Nolle said.

With movement on some of these concepts by industry standards bodies such as the IP Sphere Forum, the MultiService Forum and the TeleManagement Forum, Nolle said, “We could see worldwide trials of some of these notions in 2008. I don't think we're tremendously far away.”