Double time - U.S. Robotics' CEO Casey Cowell on 56-Kbps modem technology - Company Business and Marketing - Cover Story

Fast-talking Casey Cowell, CEO of US Robotics, is doubling modem speeds and talking about ever faster access to the public network.

For years, 9,600 b/s defined the high-end data transmission rate. Then encapsulation, compression, and advance caching techniques appeared, and data transmission rates rose into the stratosphere.

No company symbolizes the dramatic advances in modem technology more than US Robotics. And no executive articulates his vision with more enthusiasm than US Robotics' President and CEO, Casey Cowell. Cowell, a shirt-sleeve executive who still strides the corridors of US Robotics asking for opinions and feedback, co-founded US Robotics in 1976.

An industry insider frequently quoted in the The Wall Street Journal, Forbes, The New York Times and Business Week, Casey Cowell slowed down just long enough to give Communications News a snapshot of data communications as it approaches the millennium.

FASTER TO MARKET

Skokie, Illinois-based US Robotics is a leader in desktop computer modem design and manufacture. Its strategies drive the modem industry; its products give access to the Internet. They appeal to dial-up modem users, always ready for the next wave of technology.

The company commands a 40% share of the North American market. Internet service providers, online services content providers, and data center network managers all look to US Robotics for systems that allow them to manage remote data traffic.

Early in his career, Cowell saw the potential to configure DSP-based platforms as communications devices. He created a core business strategy to realize this vision, developing the firmware and software to support it. As success followed, he remained true to his vision, evolving his product line while leveraging market share and fast-to-market production techniques to stay atop the wave he had created.

A MARKET CAPTURED

"US Robotics has always been in the network access and remote communications business," says Cowell, whose delivery can only be described as rapid-fire. "In the late '80s we took our core dial-up modem technology and developed rack-mounted systems. Then we added intelligence and network management capabilities. In turn we applied this common investment to both ends of the telephone connection. This proved to place us in a very powerful position."

The company's initial focus was on client modems. "Our competitors typically acquired technology from outboard suppliers. We never did," Cowell says. "We developed our own, and set our own quality assurance standards during development. So we came to market faster, with equipment that met tough performance and price criteria.

"We competed in the desktop modem business right away, going right for the network access business. As a result, we developed very powerful and very flexible communications tools. We felt we could continue to differentiate our products from those of our competitors by continuing to develop products that went faster and maximized the transmission capabilities of copper wire."

Cowell thinks this strategy will allow US Robotics to meet the screaming bandwidth demands made by a new generation of data-intensive applications.

"We will continue to evolve the most flexible platform. At the edge of the network our focus is on two kinds of scalability: vertical and horizontal. Vertical scalability gets more ports within a given physical footprint. Horizontal scalability adds more communications technologies within the chassis while maintaining flexibility.

"Everybody understands vertical scalability; at US Robotics, we take horizontal scalability to be our imperative," Cowell says. He illustrates this point by citing the recent announcement of his company's high-speed X2 technology. This proprietary modem technology enables remote dial-up users to receive data from a network provider in the 56 kb/s range, while using the analog PSTN.

NEW AVENUES

Data exchange rates approaching 56 kb/s push the envelope for the general purpose dial-up modem, hovering at the low end of ISDN capability. Most industry analysts expect the next leap forward to incorporate asymmetrical digital subscriber line (ADSL) and very high bit rate digital subscriber line (VDSL) technologies. Cowell agrees, believing "million line ADSL markets will emerge around the millennium."

Some observers think a slow-phase ADSL introduction will allow high-speed cable modems to capture the consumer Internet access business. Cowell does not.

"ADSL and cable modems will coexist for many years. Remember that ADSL also provides switch congestion relief from all the local loop-based Internet connections. The PSTN is really only there to handle 3.0 KHz voice communications.

"Ask the question `What is access?' Thinking of access as a data pipe simplifies the equation right away. Ask the question `What else can we install?' Various shades of DSL technology come to mind. Take traditional consumer markets: multiple DSL technologies will enter customer premises to support enhanced home and small office demands."

Cowell sees analogies between the way the telcos handled ISDN (Integrated Services Digital Network) during its startup phase and the way they approach ADSL now.

"The telcos have a compelling reason to do something quickly about switch congestion. Look back a few years to the ISDN situation. Network providers couldn't make money on a line-by-line upgrade basis. Now telcos have the opportunity to configure and install ADSL as a parallel data network, eliminating the problem altogether."

If the telcos adopt a parallel data network architecture, Cowell believes several scenarios make sense: ADSL traffic could be routed through a voice switch, with data traffic routed through a different path. Both telcos and ISPs win in this scenario. If ADSL is routed instead through a parallel packet data network, more bandwidth becomes available for end-users, eliminating traffic congestion. Adopting a parallel data network architecture gives the telcos another opportunity: segmenting their market. Cowell argues continual market segmentation ought to be a cardinal rule for all carriers, producing both strategic and tactical advantages.

Strategically, segmentation allows closer market management and customized product and service offerings. Tactically, segmentation allows fast-to-market, turn-on-a-dime product and service changes, dynamically responsive to business trends and consumer response.

Convergence of networks means that networking managers can look forward to a "new generation of interconnected devices, from telephones, to videophones, to television and Internet access products, leveraging the dramatic increases in processor speed and power.

"A differentiating aspect of this new generation of products will be better integration of software and firmware," Cowell says. "Product and service upgrades can be coordinated easily from a central spot. Product performance can be coordinated within a limited distance, through power lines, telephone lines, wireless, or cable. Devices that can be used for different applications can potentially be switched over any of several different communications technologies."

Cowell imagines a near-term consumer environment in which multiprocessing is an everyday reality. "All these enhanced services will come into one central location and get distributed," he says. "The whole process could be controlled by the telephone. An intelligent switch could coordinate these services by making dynamic provisioning decisions."

For networks, says Cowell, US Robotics will "continue to do what we're good at. We'll concentrate on developing an elegant, cost-effective technology to provide faster access in a mass market environment. There's every reason to believe we can scale down our approach slightly, to cover the remote end of the network.

"We will build network access products that scale horizontally and enable multiple communications technologies to reside in one box. In the long run, we think things will end up pretty much hub-to-hub. In other words, multiple technologies from multiple networks converging on both locations."

With its new products, US Robotics "will be among the significant players in both X2 and DSL technology," says Cowell. Both technologies will be strategic elements of its overall product base and will be mass marketed, because the telcos attach a priority to solving the access congestion problem.

Cowell believes that once the mass market moves up to higher speeds using ADSL, there will be major changes in both equipment and networks.

"Suddenly we'll have these subscribers who are uploading data in the multi-kilobits-per-second range also wanting to receive downstream multimedia at multiple millions of megabits. This will place huge demands on the current switching infrastructure."

COPYRIGHT 1997 Nelson Publishing
COPYRIGHT 2004 Gale Group