We describe the design and implementation of the Distributed Threads System (DTS), a programming environment for the parallelization of irregular and highly data-dependent algorithms. DTS extends the support for fork/join parallel programming from shared memory threads to a distributed memory environment. It is currently implemented on top of PVM, adding an asynchronous RPC abstraction and turning the net into a pool of anonymous compute servers. Each node of DTS is multithreaded using the C threads interface and is thus ready to run on a multiprocessor workstation. We give performance results for a parallel implementation of the RSA cryptosystem, parallel long integer multiplication, and parallel multi-variate polynomial resultant computation.

We describe the Distributed Object-Oriented Threads System (DOTS), a programming environment designed to support object-oriented fork/join parallel programming in a heterogeneous distributed environment. A mixed network of Windows NT PC’s and UNIX workstations is transformed by DOTS into a homogeneous pool of anonymous compute servers forming together a multicomputer. DOTS is a complete redesign of the Distributed Threads System (DTS) using the object-oriented paradigm both in its internal implementation and in the programming paradigm it supports. It has been used for the parallelization of applications in the field of computer algebra and in the field of computer graphics. We also give a brief account of applications in the domain of symbolic computation that were developed using DTS. Key words: distributed threads system, heterogeneous networks, Windows NT

. This paper describes the runtime kernel of Paclib, a new system for parallel algebraic computation on shared memory computers. Paclib has been developed as a professional tool for the simple design and efficient implementation of parallel algorithms in computer algebra and related areas. It provides concurrency, shared memory communication, non-determinism, speculative parallelism, streams and a parallelized garbage collection. We explain the main design decisions as motivated by the special demands of algebraic computation and give several benchmarks that demonstrate the performance of the system. Paclib has been implemented on a Sequent Symmetry multiprocessor and is portable to other shared memory machines and workstations. 1 Introduction Computer algebra is that branch of computer science that aims to provide exact solutions of scientific problems. Research results of this area are e.g. algorithms for symbolic integration, polynomial factorization or the exact solution of algeb...

We give an introduction to programming methods, software systems, and algorithms, suitable for parallelizing Computer Algebra on modern multiprocessor workstations. As concrete examples we present multi-threaded programming and its use in the PARSAC-2 system for parallel symbolic computation, and we present some examples of parallel algorithms useful for solving systems of polynomial equations.

Abstract not found

. We have extended the task management scheme for the parallel computer algebra package PACLIB. This extension supports "virtual tasks" (tasks that are not yet executable) which are created more efficiently than "real tasks" (tasks that are immediately scheduled for execution). Virtual tasks become real only when the system is idling or existing real tasks can be recycled. Consequently, the overhead for task creation and synchronization but also the memory requirements of a parallel program may be reduced. We analyze the system theoretically and experimentally and compare it with another virtual task package. 1 Introduction The purpose of this paper is twofold: first it reports the extension of the task management scheme for a parallel programming package developed at our institute. Second it carefully investigates the semantic and performance consequences of this modification and compares them with the results reported for a system that was developed elsewhere with similar objectives...

Abstract not found

We report on the development of PARSAC-2, a library of parallel algebraic algorithms designed specifically for networks of multiprocessor workstations. PARSAC-2 is built upon the S-threads system environment for multi-threaded symbolic computation. S-threads provides virtual parallelism by mapping thousands of very light-weight processes onto the processors of a workstation. It is currently being extended with network functionality, so that heavy-weight processes can be mapped across the network while preserving the S-threads interface. The current goal of algorithm development in PARSAC is the construction of a parallel polynomial equation solver using Groebner-Bases. We report on the design of a strategycompliant parallel Groebner-Basis computation with factorization. Introduction Symbolic computation is a high-level computational task which makes it comparatively complex and slow. However, it is increasingly applied in science and engineering [FGHK94] and any significant increase i...