(Trends in future) data acquisition architectures

✍ Scribed by I. Gaines

• Publisher: Elsevier Science
• Year: 1989
• Tongue: English
• Weight: 83 KB
• Volume: 57
• Category: Article
• ISSN: 0010-4655
• DOI: 10.1016/0010-4655(89)90287-7

✦ Synopsis

A hardware overview of commercially available vector and parallel processing systems is presented, focusing throughout on the issues of good cost-performance, integration, usability and production capabilities. Following a rapid outline of the "flagship" supercomputers from Cray, ETA, Amdahl and IBM, consideration is given to the shared-memory vector processors available with the second-generation of multiprocessor minisupercomputers, typified by the systems from Alliant and Convex. We consider also the so-called "graphics supercomputers" from Stellar, Ardent, Silicon Graphics and Apollo.

Turning to large-scale parallelism, we analyse the progress made in taking such systems into production environments. Specifically we consider the iPSC/2 from Intel, the systems from NCUBE and Symult, together with transputer-based systems and the SIMD machines from AMT and Thinking Machines.