Sadik Esener and Mark Kryder
Over the last decade, areal storage densities of secondary storage systems have increased by at least 2 orders of magnitude due to the insatiable thirst exhibited by computing applications for higher capacity, faster access, higher data transfer rate and very low cost data storage systems. This extraordinary increase in areal data densities has been primarily achieved by decreasing the bit dimensions further and further. For high- performance systems, areal density remains the critical factor because mechanical and electrical constraints dictate that low cost, fast and accurate data sensing can only be achieved over short distances.
A key unknown for the future of the data storage business over the next decade is how far the applications pull will increase the areal densities. If the application pull continues, an equally important issue is whether reducing the bit size of conventional storage systems will remain the economically viable solution to increase areal densities or whether new technologies will be needed.
From the previous chapters, it has become clear that the areal density of conventional data storage systems will encounter certain physical or engineering limits within the next decade. For example, the diffraction limit of light puts a limit on the practical spot size in optical disc systems; the thermo-magnetic limit sets up a maximum areal density for longitudinal magnetic recording used in present-day hard drives.
In view of these limits it is expected that new, more unconventional technological solutions may become necessary. These solutions may entail the further increase of the areal density by surpassing the presently perceived limits, using, for example, near-field optics or perpendicular magnetic recording. The solutions may also take advantage of additional available degrees of freedom such as those provided by volumetric storage techniques. For example, holographic and two-photon optical storage approaches enable data to be recorded in the volume of a medium rather than just its surface, thereby achieving much greater volumetric densities and data locality.
The solutions may also be derived from significantly different technology platforms that may lead to new types of data storage systems. For example, micro-electro mechanical systems (MEMS) technology can enable probe storage techniques to achieve 100X better areal densities than those projected for conventional technologies. In a more distant future, biologically inspired nano-structures may be envisioned to play important roles in storage systems as well. Finally, solutions may result from indirect impact of new technology platforms. For example, the advent of low cost vertical cavity surface emitting laser arrays and of micro-mechanical actuators may significantly affect the cost and design of pick-up heads enabling fast access parallel recording and readout.
This chapter will review the possible evolution paths for alternative data storage technologies over the second part of the next decade. Along with this review we will attempt to address issues critical for devising a long term strategy for U.S. data storage systems and related component manufacturers including the following: