William D. Doyle
The information contained in this report originates from three principle sources: (1) participation in the recent tape technology roadmap program conducted by the National Storage Industry Consortium (NSIC)1,.(2) WTEC-sponsored visits to several Japanese companies including Sony, Fuji Film and Matsushita, which have a particular interest in magnetic tape, (3) contacts with JVC in Japan and Quantegy and Imation in the United States.
Magnetic tape is the oldest technology in the current storage hierarchy. It has dominated program storage and data backup in large computer systems for about 50 years, the audio market (Philips cassette) for 20 years (until the development of the CD) and the video market (VHS) for 20 years. However, the latter is threatened in the future by optical DVD. The reason for the longevity of tape in data storage is simply that it has provided the lowest cost solution for the required performance. Heads and media developed for the consumer market were incorporated into storage systems with relatively small additional effort and, because of the volume and price sensitivity of the consumer market, at relatively low cost.
The cost per gigabyte of various actual and projected storage technologies is shown as a function of time in Fig. 4.1. Traditionally, tape enjoyed a cost advantage relative to disk of two orders of magnitude, but, as the cost trend shows, this has narrowed rapidly to an order of magnitude in 1998 with a projected crossover point in less than 10 years. This is happening primarily because of the extraordinary 60%/year rate of increase in disk storage density resulting in similar decreases in cost. Tape is threatened, not by optical technology, but by disk technology unless it can sharply increase storage density at a rate comparable to disks.
Tape is a removable medium, and, consequently advances are hindered by the requirement for interchangeability, backward compatibility and agreement on specifications. There are also many formats with different tape widths and cartridge designs, which defocus development efforts. Fortunately, the major market in the future will be large and medium size systems where the number of formats is limited. The previous demand for small, low cost tape systems for data back-up in PCs, which was served by several formats, will be satisfied by optical disk or even extra hard drives. Despite these hurdles, rapid progress must be achieved, and the technology barriers to achieving it are highlighted here. Particular attention is paid to the relative strengths of Japan and the United States. We will see that Japan is in a position to control improvements in future tape systems.