To: Stephen D. Wilson who wrote (12940 ) 12/7/1999 5:24:00 PM From: Hal Campbell Respond to of 17679
Maybe CDDD are the ones to breakthrough, Steve. If it is true, buy a ton. Too many doubts here for me. This article..perhaps out of date ...summarizes the problems involved nicely. SPECIAL FEATURE: HOLOGRAPHIC STORAGE Accommodating the future deluge of data. By Kendall L. Riding The Red Herring magazine September 1998 Bandwidth. It's all about bandwidth, they say. The masses are clamoring for bigger and faster pipelines through which to send and receive an ever-increasing amount of data. But once you have a connection that lets you download the entire contents of the Library of Congress in a matter of seconds, where are you going to put all that data? Storage technology has come a long way since IBM shipped its first 5-MB computer in 1956, but there's still room for growth. A study by Japan's Optoelectronic Industry and Technology Development Association projects that by the year 2010, a storage system serving an average LAN will need a capacity of 1 to 100 terabytes, and a WAN server will require 10 TB to 1 petabyte (that's a quadrillion bytes) of storage. A consortium of corporations and universities--including IBM, Rockwell, Kodak, Optitek, Stanford University, the University of Dayton, and the University of Arizona--is working on a solution: a hologram-based storage technology. Although the result of the Holographic Data Storage System (HDSS) program has yet to emerge from the lab, scientists estimate that the technology could eventually provide a 1-TB storage capacity and data access that's ten times faster than what current systems offer--all on a crystal the size of a sugar cube. These new data storage systems use traditional holographic principles to store images of data. Binary zeros and ones are encoded and converted into an electronic bit stream, which is then recorded as a hologram on a photosensitive crystal. When the angles at which the light source enters the crystal are varied, multiple holograms can be stored on the same crystal. More than a million bits of data can fit in each hologram, and thousands of holograms could be stored on a 1-centimeter-square crystal. There are no moving parts, and because the information is read and stored in the crystal simultaneously, the result is extremely quick access. Although the idea of using crystals to store data in holographic form may sound like something out of a Star Wars movie, the principles of holographic data storage have been around for more than 30 years. IBM began researching the technology in the early '60s, but electronic components were not yet capable of encoding or retrieving the data. In the past several years, however, that's all changed. The HDSS has received a total of $50 million from private industry and the Defense Department's Advanced Research Projects Agency to develop holographic data storage technology and equipment. The final challenge for researchers is finding a viable material for the crystal, says Hans Coufal, director of holographic data storage research at IBM's Almaden Research Center and a member of the HDSS. It must be sensitive enough to record the data, but not so sensitive that any light source could erase or damage the data. The material must also be invulnerable to temperature changes, because expansion from heat could alter the crystal and destroy the stored data as well. The applications for holography will depend largely on the material that is ultimately used, according to Lambertus Hesselink, a Stanford professor and the HDSS's principal investigator. But at this point, researchers surmise that the technology may be well suited for applications with specific high-performance and large-capacity requirements, like data warehousing, data mining, and large database management systems. The estimates of when holographic applications will become commercially available range from 2003 to 2010 and beyond, so don't expect your network administrator to install a holographic storage system any time soon. And until researchers find a suitable material for the process, the technology will remain in the lab. "To solve the materials problem requires invention," says Mr. Coufal. "And an invention can't be scheduled."