The fact that tape is removable carries with it a few burdens. Purchasers of tape systems expect to be able to play their tapes on future tape storage systems, at least for a generation or two of tape drives. Thus, it is necessary for the tape drive manufacturer to not only increase the performance of future tape drive systems, but to simultaneously provide the capability of playing back tapes recorded on previous generation systems. This places very significant constraints upon the tape drive designer and has resulted in a slower rate of advance of the technology compared to disk drives. Consequently, the areal density on magnetic tape is considerably less than on magnetic disks, and thermal instabilities in the media are not looming as a short-term problem.
Rather, the problem for magnetic tape is to increase areal density at a sufficiently rapid rate to keep the cost of storage on tape well below that of magnetic disks. Until recently this was not a concern, because the cost of information storage on tape has been 10 to 100 times lower than on magnetic disks. Indeed, until recently, the major push in magnetic tape recording systems has been to improve performance (data rate and access time), rather than density. Now, however, as shown in Fig. 2.6, because the cost of storage on magnetic disk drives is coming down at a more rapid pace than that of tape drives, the cost advantage of tape is becoming smaller. If this trend continues, the time will come when tape users decide to move applications to disks. Thus, tape is now in the position that disk drives were with respect to semiconductor memory before 1991, when the disk drive industry increased the areal density growth rate to 60%/year. For magnetic tape to continue to enjoy a significant market share, it must increase areal density at a more rapid pace than it has in the recent past.
Figure 2.6. The cost in $/GB of storage as a function of the year of general availability for various storage technologies.