To get around the problems of having to sequentially access information on hundreds or thousands of feet of tape, rotating magnetic drums or disks were used. Here, the information was written to many shorter tracks where the tracks could then be accessed as required in just one rotation of the drum. With the invention and use of ferrite core memories (Figure 6) replacing the use of magnetic drums and tapes for working memory88, and transistors replacing vacuum tubes in processors and registers, the second generation of computers became smaller and faster. IBM introduced its first magnetic core memory (Figure 9) in 1954, in the 704/70589. In 1955 it introduced its first transistorised calculating machine and committed to use solid state circuitry in all new machine developments90. IBM introduced its fully transistorised 7090 in 195991. By the time of Weik's (1961a) Third Survey of Domestic Electronic Digital Computing Systems, there were more than 9000 computers in the US.
Figure 9. A 256 bit (32 byte) ferrite core memory. The donut shaped objects at the intersections of the wire are the ferrite rings (from Lee et. al, 1995).
First and second generation hardware helped corporations reduce their cost to keep and process records, but the systems were very expensive to purchase (or lease), arcane to program and labor intensive to operate (at least by today's standards)92. Given such costs, only the data processing departments of major government organizations and corporations could afford to implement them. The early computers were technologically revolutionary. However, because they were hand assembled, they remained so expensive that only large organizations could afford them. Also, as long as the clerical high-priesthood of the organizations' data processing departments owned the understanding of how to use the technology, computers had minimal impact on individual's access to knowledge or the nature of organizations (Tanner 1999; Treloar 1996; Zorn 2001; Strauss 2000).