Silicon Graphics Incorporated (SCI) was founded to deliver high-performance terminals and workstations capable of displaying 3-D objects dynamically.

By 1988, faster, second-generation workstations were introduced based on the reduced instruction set computer (RISC) architectures of Apollo, HP, IBM, Intergraph, MIPS, and Sun. Ardent and Stellar introduced graphics supercomputers that provided both very-high-performance graphics and computation capability using vector processing and multiprocessing. These entrants became the forerunners of third-generation workstations capable of providing personal supercomputing.

By 1990, HP/Apollo, IBM, SGI, and Stardent had all introduced workstations of such power that they became known as superworkstations. With the cost of 3-D graphics dropping to affordable levels, permitting use by many professionals in technical fields, these workstations are beginning to supplant traditional supercomputing.

Nearly all the traditional mainframe and minicomputer companies regard the workstation as a critical product. However, despite the great numbers of firms in the business, a scant six of them provide 95 percent (by volume) of the workstations sold. Of these, two are 1980s workstation start-ups that have remained healthy and independent: Sun and SCI. The other four are established companies that introduced workstation products to serve their customer base, which was being eroded by the new form of computing: Digital Equipment Corporation (DEC), HP (including its Apollo division), IBM, and Intergraph. These big firms intend to become even bigger players in the workstation field. IBM, for instance, has stated (in its R6000 product announcement) that it intends to capture 30 percent of the workstation market by

1993.

The rampaging progress occurring on the workstation front does not mean that the lowly PC is a thing of the past, however. Far from it! In 1990, PCs with VGA graphics have about half the resolution of workstations and only display 2-D graphics, but by using Microsoft Windows software on a PC, much of the capability of workstations (including 3-D interaction) can be provided at a low cost due to the high production volume of PCs. By the mid-1990s, it is possible that the workstation market will decline into a niche, to be replaced by a range of high-performance PCs. Alternatively, the workstation could limit the PC's upward growth. Given the difference in the software that runs on the two environments (UNIX versus Microsoft's DOS and OS/2 evolving toward UNIX), the two are unlikely to merge.

The notion of a personal workstation is a matter not just of market application need but also of technology, sociology, and economics. Owning one's own computer has been the goal of designers and users since computers were first introduced. The