Name
Abstract | ||
|---|---|---|
The increasing availability of commercial multiprocessors holds out the promise of orders of magnitude improvement in the time to solve computationally intensive problems. This promise, however, will remain unfulfilled unless we can solve some hard technical problems such as:Can we find parallel algorithms that can effectively use the available CPU cycles without getting bogged down by communication bottlenecks?Can we design adaptable multiprocessor architectures that can be programmed to solve a wide spectrum of problems at a speed comparable to specialized architectures?Can we invent effective software tools that allow us to design programs for parallel architectures and to generate efficient code for a wide variety of target machines without reprogramming?The papers by Lee and Jeong, Loganantharaj contribute to the resolution of question (1) in the areas of computational geometry and theorem proving. The paper by Shaffer investigates the implementation of a basic algorithm in real-time estimation and control applications, the Kalman filter on specialized parallel architectures. His performance results present a challenge to the viability of general architectures, as noted in question (2). The paper by Mishra and Srimani deals with an important problem in coordinating processes in a distributed environment. Their paper is directly related to question (3) since solutions to basic coordination problems are building blocks in producing efficient code for different target architectures.It is tempting to think that resource limitation will no longer be a problem with the advent of parallel processing/supercomputers. In fact, quite the opposite is true, the intellectual challenge to effectively use and coordinate multiple processors is only beginning to be appreciated. Serious investigation is needed to understand the limits of parallel processing/supercomputers. Chances are that we shall depend on them not only for processing our next pay check, but in numerous safety-critical real-time systems. |
Year | Venue | Keywords |
|---|---|---|
1987 | FJCC | basic algorithm,parallel algorithm,specialized parallel architecture,basic coordination problem,parallel processing,parallel architecture,hard technical problem,different target architecture,computationally intensive problem,efficient code |
Field | DocType | ISBN |
Distributed Computing Environment,Parallel algorithm,Computer science,Computational geometry,Automated theorem proving,Kalman filter,Multiprocessing,Software,Instruction cycle,Distributed computing | Conference | 0-8186-0811-0 |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
1 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Aloysius K. Mok | 1 | 662 | 86.52 |