Intel connected to the Internet in 1986 via a CSNET dial-up PhoneNet connection. The connectivity was justified and paid for as a research /development requirement for our engineering design and software development groups. At that time, a prerequisite for connectivity was proof that the organization was a bona fide research and development organization. By the end of 1988, our telephone charges for connectivity were just about equivalent to the cost of a leased line. By 1989, we were allocating Internet charges among three different Intel divisions, Microcomputer Development, Process Technology Development and Systems Development. During this period, Intel's internal network was evolving as four distinct networks, better connected in some areas, unconnected in others. Users demanding mail access to the Internet accelerated the deployment of mail gateways between major mail systems. Migration to (only) two mail systems was also underway. It was during this period that a survey of Intel Library users indicated a growing interest in access to information resources on the Internet....more than mail usage was on the horizon. Some internal network infrastructure was evolving in a way to make eventual information access to/from the Internet impossible. More troubling was the fact that more than one distinct internal network infrastructure was beginning to evolve. At this time, many people within Intel agreed with the view of the Internet expressed during a site visit by an Information Services person from a large US company: "...the Internet...it's pretty much still the realm of wild- eyed scientists and airhead academics". Really, something only an engineer or scientist could love.
By 1991, it was apparent that there were many network knowledgeable groups within the company; some with site or regional purview, some with corporate. There were almost as many network architectures envisioned as there were groups tasked with some aspect of network development and support. The necessity of having a common, shared vision of a Corporate networking direction became apparent. In early 1992, the Intel Libraries undertook a joint effort with users and Corporate Information Services to define a strategy and vision for the Intel Libraries. Two key aspects of this vision were delivery of information to any desktop within the company, and enhanced access from the desktop to information resources outside the company. The growing number of information resources on the Internet were of particular interest....and others besides engineers were beginning to take note.
In imagining an Intel Corporate network architecture, because Intel is a consensus-driven company, it was necessary to devise a methodology that ensured participation and buy-in from as diverse a group of network stakeholders and users as possible. Intel sought assistance in this effort from consultants who were experts in both communications technology and management process . While this was underway, an internal software summit had yielded the names of some 200 committees, individuals and groups concerned with some aspect of networking. With the help of the consultants, a process of interviews and consensus building meetings was set in motion. This culminated in a scenario planning activity which evolved the vision for the Intel Corporate network. The vision that was adopted was one of rich interoperability derived from diverse distributed systems interoperating by conforming to common standards. A subsequent meeting of all interested Intel network technologists decided that a strategy which involved evolution to an enterprise internet, based on Internet technologies, was the best way to acheive the overall vision. An overview of the resulting vision will be described.
Another area to which our Internet connectivity directly led us, was into the area of Internet usage and conduct guidelines. The results of some of these efforts have been described elsewhere (1). Our most recent effort has been in the area of Internet posting guidelines. Other current issues and concerns include broader (worldwide) Internet connectivity; redundant connectivity, user education, and marketing presence on the Internet. Our approaches to some of these will be briefly described.
In conclusion, any large company providing large scale access to the Internet for its business groups and employees is likely to see impact on the way in which it does business. Internal network infrastructure and user guidelines turned out to be two major considerations for Intel, and could also be for any large organization embarking on an Internet connectivity path.