Glenn T. Sincerbox, program manager for holographic storage systems and technology at the IBM Almaden Research Center in San Jose, Calif., passed along the following information:

"Since holographic data storage was first proposed and demonstrated in the early 1960s, many efforts have sought to make the technology more capable and affordable to computer users. We have made some dramatic progress recently, but even so, the very first commercial products are at best still three to five years away. And the wait will surely be longer for home PC users, for whom low cost is an extremely important factor.

"Holographic storage involves splitting a laser beam in two, placing a 'page' of data on one of the beams and then aiming the beams so that they cross. Where the beams intersect, an interference pattern of light and dark areas occurs. The beams are directed so that the interference pattern forms within a special optical material that reacts to light and retains the pattern. A thousand or more pages can be recorded within a single volume of material by crossing the beams at different angles, a technique called angle multiplexing. Additional volumes are created by aiming the beams at different spots within the material. Shining a 'playback' laser beam through the material (oriented at the same angle that was used for recording) produces, almost magically, a second beam that re-creates a holographic image of the original data page. That page can then be read by an array of photodetectors.

"In the past few years, there has been considerable progress in developing affordable devices that could be used to place data pages onto laser beams and to detect the illuminated holograms. Unfortunately, the best of the current holographic materials--nonlinear crystals--are very expensive and have limited capabilities. I am one of the chief principal investigators for two consortia recently funded by the Defense Department's Advanced Research Projects Agency (ARPA), along with several companies and universities. One consortium is looking for improved holographic materials; the other is developing prototype systems for specialized applications. We still have not found an all-around ideal material, although there are some promising candidates. The consortium plans to complete its first write-once data-storage demonstration platform in late 1997; a read/write (erasable) version should follow a year or so later.

"The greatest potential advantage of holographic data storage is its ability to access vast amounts of data rapidly and to transfer it to the computer's central processing unit in parallel at very high rates--billions of bits a second. Such rapid retrieval of large data pages is particularly suited for storing digital images, such as movies, medical images, maps, catalogues and so on.

"But especially in the early years of any commercial production, the general public will probably find holographic data storage systems is too bulky and expensive. Although the technology can store huge amounts of material in very small volumes of material, the entire system of lasers, optics, detectors and electronics will initially be rather large. Once the technology is successfully established in one or more business niches, additional technical and manufacturing development, along with economies of scale, could help make holographic devices more affordable and compact. But popular applications of this technology are at least several years away."

Lambertus Hesselink of Stanford University, one of the leaders in the field, offers some additional information:

"Holographic data storage is receiving a lot of attention lately, because there is a strong need for large data storage systems having capacities exceeding hundreds of gigabits and fast access times as well as fast data transfer rates. Currently, only RAID systems (which incorporate two or more magnetic hard disks for storage and backup) can achieve this performance, but only at a very high cost, as many drives need to be connected together. Because the medium is not removable, capacity can be increased only by adding more drives to the system. Other systems also have serious drawbacks. Optical-disk 'jukeboxes' are clumsy and mechanically unreliable. Tape drives do not afford efficient, random access to the data. In the future, systems built around high-density digital videodisks (DVDs) will have very large data capacities, but their access times and data transfer rates are limited

"Holographic systems are attractive because they can store very large amounts of data inside a volume of recording material. The surface data density in some of these systems exceeds 100 bits per square micron, or 10 billion bits per square centimeter. The primary components of such a system are a spatial-light modulator (SLM), which formats the data to be stored into a 'page' of pixels; an array of charge-coupled devices (CCDs), which 'read' the data stored in the hologram; and the recording medium itself. A laser is used to record and to retrieve the data.

"The sensitivity of the recording medium usually dictates the type of laser that is required. Currently an yttrium-aluminum-garnet laser is the most commonly used illumination source, but we would like to use a smaller, cheaper diode laser that emits in the near infrared (at wavelengths just too long to be visible to the eye). Unfortunately, most traditional ferroelectric recording materials are not sensitive to those wavelengths. Recently, however, the government-industry photorefractive information recording materials team (of which I am the principal investigator) has made substantial progress in the development of materials that are sensitive in the near infrared. Holographic images can be 'fixed' into these materials using lower-power lasers than has been the case in the past. In addition, new polymer materials are currently under development; these promise to be cheap and very sensitive, which again allows the use of much lower-power lasers.

"These new components have been incorporated into promising holographic data storage systems. After the first digital holographic data storage system was built at Stanford University (see 'Volume Holographic Storage and Retrieval of Digital Data,' by J. F. Heanue, M. C. Bashaw and L. Hesselink in Science, Vol. 265, page 749; August 5, 1994), several groups around the world have built similar systems. The storage capacity and data transfer rates of these devices have improved significantly since then.

"In summary, we have made a lot of progress in the past few years, but the recording materials are still the key component limiting the overall performance of holographic data storage systems. Much work is currently being carried out at universities and industrial labs around the world. Based on the progress in those labs, it is reasonable to expect that viable products could reach the market around the turn of the century."