Timeless thoughts on Timesharing

This post is NOT about vacation homes!

I recently attended an event at the Computer History Museum titled “Timesharing / Remote Processing Services”.  Though the gathering was intended as a look back at nearly half a century ago, I was fascinated by how many of the concepts were very much still at work in contemporary computing, albeit in different configurations.

The conference brought together 15 or so executives who had founded and/or run companies in the 1960s and 1970s that most of you have never heard of, including Comshare, Compuserve, Tymeshare, NCSS, GEISCO, IDC, On-line business systems, and my company, Ross Systems.  My experiences at Ross Systems were part of the catalyst for the formation of ExpertCEO.

Wikipedia defines ‘Time-sharing’ as a computing resource shared among many users by multitasking. Its introduction in the 1960s, and emergence as the prominent model of computing in the 1970s, represented a major historical shift in the history of computing from “batch processing” to interactive (“on line”) computing. By allowing a large number of users to interact simultaneously on a single computer, time-sharing dramatically lowered the cost of providing computing to the individual, while at the same time making the computing experience much more interactive.

During the two-day session, attendees discussed a number of topics, including technology (and its constantly decreasing cost), business models, applications, communications, and industry consolidation. The similarities between today’s computing paradigms and issues and those that existed in the 60s and 70s continue to amaze me.

Two important trends today, virtualization and cloud computing almost perfectly parallel the timesharing concepts, albeit with vastly differing technologies and economics.  However, the basic concepts of better machine utilization, sharing of centralized computing resources and a better user experience are common to both eras.

Virtualization – Refers to the technology to divide one physical machine into many “virtual machines” so that multiple, independent programs can run simultaneously.  The primary benefits are better machine utilization, server consolidation, reduced energy costs, etc.

Cloud computing – A model of computing in which dynamically scalable and often virtualized resources are provided as a service over the Internet.  Users need not have knowledge of, expertise in, or control over the technology infrastructure in the “cloud” that supports them.   Software as a Service (Saas) can be considered part of cloud computing (see my blog post titled “The Future of SaaS”).

Some differences between then and now are worth noting. Communications speeds when time-sharing was prevalant began at 11 characters per second (cps) with a teletype and gradually grew to 30 cps and then 120 cps.  Compare this with today, where my Comcast cable modem operates at 6 million bps or about 600,000 cps.  An 88 MB (that’s an “M”) disc cost approximately $40,000 in 1975 or about $450 million for a TB of storage.  Today, you can go to Fry’s and purchase a 1TB disk for a couple of hundred dollars.  This price decrease closely approximates Moore’s law which indicates that the improvement in technology will double every 2 years.

What was the primary cause of the demise of the timesharing industry?  The dramatic improvements in technology that continue today were also occurring in 1981.  IBM released its first PC in 1981. PCs were very inexpensive, offered a better user experience, and there were numerous apps available at a relatively low cost (e.g. Lotus 1-2-3).  The cost and complexity of the infrastructure associated with a central timesharing computer could no longer compete with the lower cost and better interactive experience of PCs for these types of applications.

Today, Microsoft, Google, Amazon and others are building huge, complex data centers to support cloud computing infrastructures and SaaS apps.  There’s an interesting article in the June 8 edition of the New York Times Magazine titled “Data Center Overload”.  It is a fascinating description of the size and complexity of the data centers that are now under construction, and the energy costs that are required to power and cool them.

Babbage Engine - Computer History Museum

Babbage Engine - Computer History Museum

The timesharing industry had a great ride for about 10 years.  Ten years from now, might the same trends that doomed timesharing, cause the demise of the cloud computing infrastructure?

A final note: The computer history museum contains a wealth of information and very interesting exhibits.  I encourage people to visit the museum and give to its support.

Leave a Reply