This paper discusses the digital computer revolution that started two decades ago, and outlines some of the economic aspects of computer utilization. It summarizes areas of past and current computer applications, and indicates the areas where problems may be solved by extension of present computer technology and development of new methods. It discusses some of the technical training revisions being made to take advantage of present and future machine capabilities. It also emphasizes the increased speed, reduction in cost and profitability offered by today's machines. Managers and engineers must be aware of the economic usefulness of today's digital computer.
Today we stand at the threshold of a new era in science. Refineries of today make the refineries of the 1930's uneconomical and obsolete. Transportation systems, rigs, completion techniques, reservoir engineering methods, production practices. etc., have advanced far beyond the imagination of most early petroleum pioneers. One seemingly unrelated tool has caused another revolution in the entire petroleum industry, and especially in the area of economic planning . . . this tool is the electronic digital computer.
In view of the fact that modem computer history covers a mere 20 years one would be inclined to think that very little history could be written. Yet, life the primary function of the computer itself, a great deal has happened in a short time. The Association, for Computing Machinery recently referenced and cross-catalogued about 5,000 articles, books and papers covering computing science literature written during the four year period from 1960 through 1963.
From 1944 to 1965 newer, better, faster and more compact machines have been developed in such rapid succession that only computer specialists are able to keep up with the advances. Instead of reaching what might be considered a plateau, these improvements have only pushed the horizons further and further aside, opening ever wider areas of potential development.
A small computer might perform 150 additions in one second, which would seem like a respectable speed. but large machines currently available can perform an amazing three million such additions in one second. There is a rental price differential between the large and small machines of about 200 to 1; some of the faster machines can produce more than 100 times as many computations per dollar's worth of operating time as the smaller ones.
It has been estimated that using an IBM 7094, for some applications, that at a rental charge of $575 an hour, one dollar can buy as much calculating as one man could perform in a year. This, of course, is based only on actual machine time used and does not include the cost of program development. Computers even faster than the 7094 are currently in use, and newer machines can calculate about 20 million times faster than humans and with fantastic accuracy. Because of this, problems can now be solved that could not even be considered 20 years ago.
Although progress in computer development has advanced rapidly, one large obstacle to even greater computer use has been the tedious preparation of coded instructions for the machine the program. However, progress has been and is being made in the field of master programs (assemblers or compilers) which in the future may be able to take a problem posed by the operator, work it over, figure out how to solve it, and then either give back the answer or tell what data might be missing.
Approximately five billion dollars was spent in the United States during 1964 for computers and associated electronic processing systems operated by business, government, and the military. A multibillion dollar a year foreign market is expanding at the rate of 25 per cent annually.