Introduction
Computer tape systems store data using magnetic recording technology. Demand for ever-increasing backup and archival storage capacity has led to very high recording densities, making reliable extraction of data stored on the magnetic media a very challenging task.
Since the announcement of IBM's first tape drive—the IBM 726 tape unit—in 1952, tremendous progress has been made in tape storage technology, see Figure 1. The IBM 726 magnetic tape unit was able to store data at 6.1 kB/s at an areal density of 1400 bit/in2 and had a capacity of approximately 2.3 MB. In 2008, IBM announced the TS1130 tape drive with an uncompressed capacity of 1 TB in a single tape cartridge and a native data transfer rate of up to 160 MB/s. This translates into an increase in data rate of roughly 4 orders of magnitude and more than 5 orders of magnitude in areal density.
In order to remain the medium of choice for tertiary bulk storage in enterprise systems, tape-based storage needs to match or exceed the performance growth of hard-disk systems, while maintaining its price/GByte advantage. Fortunately, there is still significant potential to continue scaling tape technology: In 2006, IBM demonstrated the feasibility of pushing the storage capacity of a tape cartridge to 8 TB. A recent feasibility study indicated that there is potential to continue scaling tape to densities towards 100 Gb/in2 [1], [2].
Maintaining such areal density and data rate growth poses complex technical challenges. In order to increase tape cartridge capacity while maintaining the same cartridge form factor, the tape area needs to be increased and/or data needs to be written spatially closer together. The area available for storing user data can be increased by improving the format efficiency or by making the tape longer, which requires the use of thinner tape. This in turn renders the tape more fragile and necessitates advances in the tape transport system and tape path design. Capacity increases can also be achieved by scaling the linear density or the track density. While continued incremental increases in linear density can be expected in the future, there is a much larger potential for increasing track density by reducing the track width. In order to exploit this potential one has to minimize the tape lateral motion and compensate for any remaining disturbances by adjusting the heads dynamically to the varying track positions. For very high track densities, positioning control down to the nanometer scale will be required. Such precise control necessitates significant improvements in the tape media dimensional stability, as well as in the performance of the servo pattern and servo channel, the actuator performance and the track-follow servo controller [1], [3], [4]
The "read channel" illustrated in Figure 2 processes the signals from the servo and data readers and extracts the stored information. It must be designed to deal with a large amount of variability that results from cartridge exchange, backward compatibility requirements, and aging effects [6], [8]. The data channel, which is illustrated in Figure 3, is responsible for the user data detection. The reduction of the bit size associated with areal density increases poses a major challenge for data detection due to the resulting decrease in signal-to-noise ratio (SNR). Significant improvements in the signal processing techniques to recover the data from the tape are required in order to deal with this loss of SNR and to meet the demand for ever increasing data rates and the requirement of media interchangeability.
Our work in tape storage focuses on
- devising novel signal processing and coding techniques,
- introducing very precise and robust tape transport mechanisms and track-following servo systems,
- optimizing the tape path design in order to guarantee very high reliability of IBM's tape drive products operating at ultrahigh recording densities, and
- designing novel heads and actuators.
