Abstract

Optical memory systems utilizing volume holographic storage (VHS) have received considerable attention in recent years^[1-3]. One aspect of VHS is the ability to copy stored data from an archive (a VHS medium containing N permanently-stored data sets) to a secondary storage medium (SSM). To achieve the copying of volume holographic memories from the archive to the SSM, three primary approaches can be utilized: parallel^[4], incremental^[5], and sequential^[6]. In the parallel approach, there exist N mutually incoherent read beams that enter the archive and read out and copy all N stored data sets simultaneously. In the incremental approach, there exists a single coherent read beam that is rapidly multiplexed over time (in angle, wavelength, phase, et cetera) in a manner such that it reads out and copies all N data sets serially, over many repeated iterations in which each iteration time is short compared to the SSM’s response time. In the sequential approach, there also exists a single coherent read beam that is multiplexed over time (in angle, wavelength, phase, et cetera) but such that it reads out and copies the N data sets sequentially in a manner that follows an appropriate (single-pass) exposure schedule. In this paper, we explore the fundamental material limitations in each of these copying approaches for the cases of all-optical, quasi all-optical, and hybrid opto-electronic copying schemes. These limitations include the maximum allowable intensity, I_A, into the archive that will not damage it, the resulting total intensity, I_D, of the diffracted beam from the archive as a function of N, I_A, and other archival material and geometry constraints, dark conductivity, shot and Kerr noise in the SSM, cross-talk, and the characteristics of any devices that may be placed between the archive and the SSM.

© 1995 Optical Society of America