Extreme's Crystal Balls
In a speech called "Is There Anything Left to Invent?" that was big on vision (but short on talk about new products from Extreme -- see Extreme Intros 10-GigE Platform), Stitt told the crowd that down times are the right times for companies to innovate: "From my perspective, lower demand from customers means we can apply our resources to fundamental R&D."
In fact, Stitt almost seemed to be saying that it is the duty of companies to keep developing "disruptive technology" rather than just focusing on product upgrades. "I think that as an industry we are betraying your trust if we get too caught up in providing incremental improvements."
Stitt predicted that over the next three to five years there will be three major factors that transform enterprise and, to a certain degree, carrier networking. Namely: a massive amount of new devices that connect to the network; voice, video, and data applications; and ensuring that the network can scale to meet business needs.
The proliferation of devices linking with the corporate network won't just include new wireless computing and communications devices, according to the Extreme Big Cheese, but also "non-computing devices." Kit that comes under that heading includes temperature control systems for offices and robotic manufacturing systems.
The network of the future, as Stitt sees it, will also need to be able handle a greater volume of "latency sensitive" voice, video, and data apps. "Are you prepared to build a separate corporate network for each emerging application? I don't think so."
Yet the integrated network of tomorrow will also need to adjust to changing business needs, the good times and the bad, Stitt warned. "You need a network that will work on a least-cost basis, a bang-for-the-buck basis, and an all-cannons-blazing, go-for-broke basis."
So, what would one call such a wondrous network? "You're looking for the first universal infrastructure, and that is the perspective that the industry needs to take, which will require a hell of a lot of innovation."
A casual observer might think that some of the startups working in the enterprise wireless field today are taking some of the first steps down that path.
Yet, despite glancing back at Extreme's own start in Ethernet networking in 1996, Stitt had some harsh words for the startups working on corporate wireless LAN technology, a market his own company has just entered (see Extreme Switches Both Ways). Stitt made several generalized jabs at wireless startups working on technology to enable 802.11 in the enterprise, saying that many of them do not have "the scale or the perspective" to deal effectively with marrying wired and wireless corporate networks.
"You may be doomed if you go with them," the Oracle (not affiliated with Oracle Corp.) intoned. "If they call themselves a wireless company, they haven't addressed the challenges of extending the enterprise.
"The big idea out there today isn't wireless." The real challenge, continued Stitt (who was clearly on a roll), is to incorporate wireless features as part of an entire network controlled by a single console.
This won't come as news to many of the wireless LAN switch startups in the market now, as they frequently play up the ability to extend wired networking features into the 802.11 domain.
However, Keerti Melkote, a founder at Aruba Networks Inc., laughed off suggestions that the "wireless" tag might be the kiss-of-doom in this market. "Of course we're a wireless company, we're a wireless networking company," he snorted.
— Dan Jones, Senior Editor, Unstrung