Changes
By the late 1980s, hard disk drives (invented by IBM and debuted in 1956) had undergone dramatic increases in capacity and areal density at the same time cost-per-gigabyte dropped steeply. Storage innovations such as arrays of disks, optical devices and other technologies entered the marketplace and threatened to relegate tape to nothing more than backup or archival repositories. The IT industry predicted "the death of the mainframe" as well as the demise of tape storage.
Dr. Paul Low, president of IBM's General Products Division, and Juri Matisoo, director of the Almaden Research Center, convened a group early in 1988 to examine the viability of tape as a storage medium for the future. One member of that task force, Dr. James Eaton, recalled, "Early in the study it became apparent that we had to make radical improvements in linear tape system density and performance. Dramatic improvements were clearly possible. Tape technologies were available with over 100 times the density of the very successful IBM 3480/3490 family of tape devices." Although such existing tape technologies improved capacity, they did not offer the data rate required for the applications that IBM storage products were focused on. The goal was to achieve both increased capacity and data rate without sacrificing the proven reliability of the 3480/3490 products that had become a world standard for information storage on tape.
At the conclusion of the task force, the prevailing attitude among the group members was one of excitement, with a reinvigorated focus on advancing IBM's presence in tape technology. Dr. Eaton summarized, "At the end of the task force study, it was obvious that data storage tape areal density could be increased by hundreds of times. We had to get busy to make it happen or risk losing the business to whoever did!" Thus began the "New Tape Program' -- or NTP -- to explore novel directions for tape. At the same time, enhancements to the IBM 3480/3490
enterprise class of tape devices continued.
Management of multiple tape formats and architectures became an increasing problem for large and midrange customers throughout the 1990s. IBM combined the strengths of its world-class tape drives and control systems with the advanced connectivity of ESCON and then fiber support into the 3495 Tape Library Dataserver, introduced in 1993. With this large automated tape library, the IBM Tape Development Laboratory reentered the automation arena it had begun nearly two decades before with the 3850 Mass Storage Subsystem. A more efficient and higher capacity tape library dataserver, the 3494, followed shortly thereafter. Improved management of storage for disk and tape along with innovative cache-control systems and software accompanied these devices.
IBM 3494 Tape Library Dataserver
Once again, the CPU and, by that time networks of CPUs, were freed from managing storage. Control of tape volumes migrated to increasingly powerful library managers that were essentially "computers" residing in the library itself for the express purpose of managing the in-and-out flow of data. By the 1990s, the amount of data in a tape library could be hundreds of terabytes; by the end of the decade -- petabytes.
Simultaneous with rethinking tape technology, IBM underwent a significant change at its highest level. Louis V. Gerstner, formerly of RJR Nabisco and American Express, assumed leadership and immediately shifted the energies and focus of the company. Soon the tape development team in Tucson, Arizona, felt the sense of urgency generated from this new CEO, who had not grown up in IBM and who brought a fresh, external perspective to the corporation. Kevin Reardon, who headed the restructured tape team in 1995, recalls that time in the history of IBM Storage. "During this period, which eventually became recognized as the most significant and dynamic time of change in IBM's history, this team resurrected and forever changed the course of magnetic tape storage." |
