Magnetic tape plays a crucial role in assuring data integrity and is the lowest cost form of data storage. Wherever data integrity cannot be compromised, as in many forms of financial data, tape is used.
Working with our media partners, research in this area is focused on technologies for substantially increasing recording areal density to continue to enhance the value proposition for tape. Work at the Almaden Research Center is centered on exploring new path mechanics for significantly increasing track density, investigating head-tape tribology at higher linear densities, and developing a detailed understanding of magnetic recording on particulate and film media.
Part of our efforts has been directed toward establishing areal density demonstrations. In 2003, we wrote a full cartridge at an areal density of 1 Gbit/in2 (http://www.research.ibm.com/journal/rd/474/childers.html) and in 2006, we showed that with a new media and giant magnetoresistive heads, an areal density of 6.7 Gbits/in2 is achievable. (http://www.insic.org/annualmeeting/INSIC_2006_Symposium_Agenda.htm).
The recording areal density is critical to the viability of tape as a low-cost archival/back-up and restore data-storage system. While tape still is the low-cost form of magnetic storage, the large disparity in recording areal densities between hard-disk drives and tape threatens this key distinction of digital tape storage. Fig. 1 plots the areal densities of hard-disk drives (HDD) and tape drives as a function of the year of introduction. The squares represent HDD products, with a line drawn to indicate the aggregate average rate of areal density growth. While historically, this rate of areal density growth has been as high as 100% per annum, it has slowed in the past few years. The circles represent linear tape products, which show a rate of areal density growth of 42% in the past few years, now comparable with the hard-disk product trend. The diamonds indicate the areal density of helical scan tape drives, which are higher in areal density than linear tape drives. It is evident from this chart that there is a large difference between the areal density in HDD's and that in tape drives.
Decreasing the recording areal density gap is critical to the future cost-effectiveness of tape, although it can be argued that a finite gap in areal density between HDDs and tape systems is desirable and is an important aspect of the value proposition of tape. A sustainable difference in areal density between HDDs and tape products implies a time lag in the introduction of magnetic recording technology for tape products. This in turn results in a reduced required investment in technology or a leveraging of the investment already made for hard disk drives.