Publisher's Synopsis
It is becoming widely accepted that parallel computing is the only way to achieve the processing speeds that are required to meet the future needs of computer users. However designing and building larage parallel computers may not be the most difficult task: there are difficult software problems too that require the development of new solution algorithms which are capable of supporting many parallels. The emphasis of this conference was on the applications of parallel processing, on the implementation of specific applications on parallel computers or with the development of efficient mathematical algorithms to be used in applications.;Papers by Paul Messina and Jack Dongarra (University of Tennessee) provide an overview of the basic issues and trends in hardware systems software development. Jeremy du Croz (NAG, UK) talks abut the eveolution of parallel algorithms in dense linear algebra. Mike Brady et al (Oxford) write on parallel algorithms in vision and robotics and R.H. Bisseling (Utrecht) on parallel iterative solution of sparse linear systems on a transputer network.
