|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Four Decades of Modern Computing: A Retrospective for the Future
Part 1 of a two-part seriesBy Bill Howard, Former Chief CIO Advocate and Advisor to the CEOThe third quarter of 2005 is about to close, and with it I am concluding my 42nd year in the world of computers. For the last 21 years, I have led IT organizations at three large global companies. At the end of September, I will retire from Sun to spend more time with my family and to pursue interests that have been mostly "on hold" during my tenure in IT. This two-part series will be my final editorial as executive sponsor for the Sun Inner Circle program. I would like to share with you some observations from my long career in information technology. I will wrap up Part 1 by identifying a few growing problems the IT industry must address. In Part 2 next month, I will provide some predictions for the future of the industry. To see the future more clearly, I have always found it helpful to understand the past. A quick retrospective will help set the groundwork for some predictions about the future of IT. It has been a great ride working with bright, dedicated people to apply computer and communications technology to challenging global business problems. The first observation I would like to share is this: In four decades of my professional life, each phase of modern computing has looked new, but at a deeper level, much has stayed the same. When I started in computers in 1963, the industry was young, brash, and innovative, growing at 15 to 20 percent year. Although industry growth has slowed, there are still big opportunities ahead for those who innovate in technology, develop new business models, or deliver quality services. It is impossible to cover all the dimensions and nuances of the industry in a short letter, so I will attempt to capture the essence of each computing cycle, or era, from a practitioner's perspective. Let's look back at the major innovation waves for business computing: What were the major trends and milestone events? The waves of time for the modern computer era are roughly defined as generations based on the underlying technology:
ENIAC, Univac, and the Dawn of an Era The use of machines to do calculations has a rich history going back at least two millennia. But it was not until the early 1950s that programmable computers for general purpose computing started to be widely adopted in large business enterprises. It is generally accepted that the first general purpose programmable electronic computer was the ENIAC which appeared in late 1945 or early 1946. It was used for ballistics and analysis for atomic energy. The business equivalent that followed was called the Univac and was delivered to the U.S. Bureau of Census in early 1951. This computer ushered in the general purpose commercial computing era. These early computers were big and complex. The people using them were generally brilliant scientists and engineers. Security and support was relatively easy since only a few people were involved with building, operating, and maintaining the systems. Big Iron, Big Business Throughout the 1950s and 1960s, the predominant hardware platform was the mainframe -- also known as "big iron." Many manufacturers from the U.S., Europe, and Japan were in the general purpose computer business but the dominant players were known as "IBM and the Seven Dwarfs," or later as IBM and the "BUNCH" (from their initials, Burroughs, UNIVAC, NCR, CDC, Honeywell). IBM gained an early market advantage as a leader in business computing from its strong position in the electronic accounting machine business. From its stronghold of punched card machines in the accounting departments of large corporations, IBM was able to quickly branch out. IBM's work on the SAGE Air Defense System and later on the American Airlines SABRE System gave it another edge in the rapidly growing general purpose computing business. As the industry grew, it became necessary to hire and train an army of people to sell and support the computer. Since there were no formal college programs, the selection and training of people to work in the industry was mostly the responsibility of the manufacturers. Recruits had to take a test called the PAT (Programmer Aptitude Test). Successful recruits generally came from the ranks of scientists, engineers, mathematicians, linguists, and musicians -- those who were adept at symbolic representations and/or quantitative reasoning. Security and support, while increasingly difficult, was still manageable because the user population was relatively small and few people had the time or inclination to intentionally disrupt the systems. By 1964, IBM introduced the System/360 family of computers — the result of a huge "bet the company" gamble that set the standard for the rest of the mainframe era. At the time of the S/360 announcement, I was a trainee at IBM and helped with the logistics for the big customer event in the San Francisco Bay Area. The S/360 was based on the integrated circuit and allowed further miniaturization of the computer. Timesharing and Minicomputers During the 1960s, timesharing was initially adopted by engineers, scientists, and financial analysts. Some broader uses were tried by other early adopters, but the emergence of minicomputers effectively ended the timeshare era because of high communications cost, limited capability of terminal devices, and slow line speeds. The seeds of a minicomputer revolution were planted as early as 1960, with the introduction by Digital Equipment Corporation of the transistor-based PDP1 for technical computing by engineers and scientists. The PDP1 was succeeded by the PDP5, then the highly successful PDP8 in 1965 and the PDP11 in 1969. Departments were already starting to seek autonomy from the highly centralized and expensive mainframe organizations (EDP and MIS departments). Engineering, manufacturing, and scientific departments were starting to take note of fast, inexpensive computers that could be dedicated to single-use organizations. With the arrival of the integrated circuit, a new group of competitors appeared on the scene. More than 100 companies were in the minicomputer fray by the early 1970s. Minicomputers continued to grow in popularity through the 1970s and early 1980s, attacking the large installed base of mainframes and also opening up new application areas for shared computers, such as word processing. During the minicomputer era, schools and universities started to provide a broad range of curriculum and degrees related to the computer industry to keep pace with the rapid expansion of the industry. Security and support challenges, while greater because of the broader use of computers, were nonetheless still manageable. The techniques were basically extensions and refinements from the mainframe experience. The minicomputer revolution suffered much the same fate as the mainframe business when it continued to hold on to successful business models too long or failed entirely to see the power of the next cycle of computing — the PC era. Ken Olsen, the CEO of Digital Equipment, was reported to have said at the World Future Society, "There is no reason for any individual to have a computer in his home." PC Era and Networked Computing By the late 1970s, the enterprise user community was again restive and wanting to escape from the centralized high cost "mainframe" or "departmental computer" that needed to be shared with others. With the advent of the microprocessor, the PC, and local area networks, each user could have his or her own computer. This trend was reinforced by the fact that a computer at home linked the workplace and the home as never before. The advent of the microprocessor also ushered in the networked workstation. Sun Microsystems was founded in 1982 to provide networked high-performance technical workstations based on open standards. Sun's motto for the last 23 years has been simple yet powerful: "The Network is the Computer." As the popularity of PCs and workstations grew exponentially, the challenge of training and supporting and providing users with a secure computing environment became a huge issue. No longer were there a few million users, now there were 100s of millions — some very sophisticated, but many more untrained and naive. Both ends of the user spectrum present big challenges for manufacturers, service providers, and the users of computers themselves. On the highly sophisticated end, we see viruses, worms, and identity theft, while from the other end of the spectrum there is demand to simplify the user experience. The era of the microprocessor and the Internet (which really began in the late 1960s and early 1970s, but exploded in the 1980s and 1990s) has been well documented in the popular press and cannot be adequately covered in this short letter. We are now at the point where a server on a chip is a reality, software from the open source community is rapidly being adopted, and communication costs continue to drop while bandwidth increases. Also, the lines between work and home life are continuing to blur and the technology tools needed to operate in either environment are much the same today. Despite Successes, Much Remains to Be Done Most of this letter has been about the evolution of computers from a hardware perspective. It would be fun to review the other dimensions of information technology in the same way: The telecommunications impact on computers and vice versa; software evolution from machine language to Java Web services; the progression from custom built, to buying individual applications, to buying a suite, to buying software as a service; the changing role of the IT professional in the enterprise; and the promise of the Internet to bridge the digital divide. Much has been accomplished by the industry participants, but much more remains to be done. From the dawn of the information age in the early 1950s, visionaries saw enterprise computing as a totally integrated management information system. The promise of each wave of technology was that executives, managers, and white collar and blue collar workers alike would have near real time information available to do their jobs in a highly efficient manner. We are far from achieving this vision after 50 years of progress and experimentation. In the enterprise, too often I hear the lament that "I am buried in data but don't have the information I need to do my job effectively." There are still technology challenges in terms of ease of use, reliability, and quality that need to be addressed. But internal political and organizational issues remain a bigger challenge to achieving the vision of a totally integrated system that delivers the right information, not just data (e.g., the balance between centralization and decentralization; autonomy and cross organizational cooperation and collaboration; creativity and standards/processes to name a few). The pioneers of this industry have done much to set the stage for continued breakthroughs in the delivery and use of technology. Those who continue this work need to study the successes and failures of this young industry to realize the early vision that has not yet been delivered. Looking to the Future For IT professionals, although each passing phase may have seemed to be a "new, new thing," the fundamentals have remained the same. Technology has rarely been the essence of a problem or of the solution. Harkening back to one of my earlier newsletters, I call your attention to the formula Q*A=E. That is:
This was true in 1964 when I was a trainee at IBM, and was just as true in 2005 when I decided to retire from Sun Microsystems. The human and business leadership qualities that made an effective IT leader in the 1960s are very much the same today. So, I close with a few things that I feel are top priority for the industry to address to improve the quality of the user experience:
These issues are primarily addressed to the providers of information technology for millions of consumers, but also apply to business computing. As mentioned earlier, the lines between personal life and business life are blurring at a fast rate and the technologies supporting both are looking more alike every day. Join me next month for the second part of this letter which will make a few predictions about the future based on observations of the past. Bill Howard |
| |||||||||||||||||||||||||||||
| | |||||||||||||||||||||||||||||