In the late-1980s, IBM helped create a network of supercomputer centers dubbed NSFNET (the National Science Foundation Network), one of the first networks to use TCP/IP. The project essentially gave birth to the Internet—and business and life around the world changed forever. Before the Internet, scientists and researchers had to travel—often out of the country—for computing resources and to collaborate on major projects. By the early 1980s, an early Internet had begun to emerge: a primitive, regional telecommunications network linking several national laboratories and supercomputing centers that could be accessed only by trained experts. It was complicated, unfriendly and slow. But it was an important first step to the worldwide establishment of the Internet.
In 1985, the National Science Foundation (NSF) launched an initiative to build a state-of-the-art national backbone network, an inter-net, that would be based on transmission control protocol/Internet protocol (TCP/IP) and would link supercomputer centers and regional academic networks. TCP/IP is the telecommunications protocol framework developed by the U.S. Department of Defense in the 1970s, which became such a crucial part of the Internet’s plumbing. A number of universities and companies participated in its development, including IBM. Recognizing that much of this new network would have to be invented and lashed together, the NSF solicited proposals and awarded the project to IBM, MCI, the State of Michigan (home to a large community of computer scientists and keen to link up existing telecommunications networks within the state) and a consortium of universities in November 1987.
In the beginning, it was not clear if the NSFNET project could even be done. The award was met by the scientific community with skepticism, according to Hans-Werner Braun, co-principal of the project. The conventional thinking was that the task was technically undoable. But discussions among academics, government and industry began and IBMers were keenly interested. Al Weis and Barry Appleman from IBM Research and Bob Mazza, Walter Wiebe and Rick Boivie from IBM’s Academic Information Systems Division quickly engaged, offering hardware, software and project management. Senior leadership in the Research Division recognized the benefits of such a project: it was important to the nation, innovative technologies would be created and shared among the participants, and new business possibilities loomed—if this could be pulled off. Al Weis explained that before NSFNET, “IBM was unable to interconnect its large mainframes and some of its new workstations to all the research communities’ networks,” but this project held out the promise of fixing that problem.
The NSF liked the plan and authorized funding. IBM’s project manager, Harvey Fraser, recalled that “the team worked well because we had resources, executive time and the desire to make it a success.” In addition to creative attitudes, teaming and technologies, IBM, MCI and others brought the disciplines of business processes and project management to the effort, and a willingness on the part of everyone to work long hours. Some IBMers worked 100-hour weeks for months at a time.
IBM assembled a team from across the company, led initially by Jack Drescher, from Research Triangle Park Laboratory. They came up with the notion of creating a depot or assembly line, in the manner of Henry Ford, and a skunkworks atmosphere developed in Michigan where the project was headquartered. Equipment and parts flowed in from IBM and other companies, such as computers and peripheral gear, which they configured and tested then deployed to various campuses and super computer sites. "We turned the third floor of the Computer Center into an assembly line.… In the end we had the whole floor covered with parts and machines and boxes; it was a great way to deploy everything," remembers Elise Gerich who was the site liaison at the time. In fact, 150 systems with thousands of machines, parts and telco equipment were implemented. Eight months later, in July 1988, the network went live with 170 networks linked together, making it possible for the first time for the academic and research communities to access a high-speed, reliable and effective data network service that spanned the United States. The previous network, ARAPANET, was shut down soon after, since the new one worked so well.
Capacity became a problem almost immediately, requiring IBM and its partners to continue delivering more innovative technologies and equipment throughout 1989 and beyond. In IBM’s case, this often involved incorporating new networking capabilities into various software products, a process that led to the development of Internet-related software and consulting offerings over the next twenty years. In fact, that year alone traffic grew by 500 percent, beyond everyone’s wildest expectations. The big next step was moving the network to a T3 capable backbone (a new generation of higher speed digital switching) and along the way, using an IBM RS/6000 for each T3 node.
Traffic volume grew and grew. Back in 1988, only users in the US, France and Canada accessed the network. Between 1989 and 1993, ten to twelve additional countries were added each year, 21 in 1994 alone. By the time the transition was made from the NSFNET to the networks of the various telecom carriers and commercial Internet Service Providers in 1995, the network reached more than 50 million users in 93 countries. The network accelerated its response time and expanded capacity and functions as it migrated from T1 to T3 speeds and technologies. The project brought IBM squarely into the world of the Internet, exposing scientists, researchers, product developers and field organizations to new technologies and innovative uses of telecommunications. The NSF considers it one of its most valuable contributions to the nation: the forerunner of the modern Internet.