An automatic parallelization framework for algebraic computation systems

Abstract

This paper proposes a non-intrusive automatic parallelization framework for typeful and property-aware computer algebra systems. Automatic parallelization remains a promising computer program transformation for exploiting ubiquitous concurrency facilities available in modern computers. The framework uses semantics-based static analysis to extract reductions in library components based on algebraic properties. An early implementation shows up to 5 times speed-up for library functions and homotopy-based polynomial system solver. The general framework is applicable to algebraic computation systems and programming languages with advanced type systems that support user-defined axioms or annotation systems.

An automatic parallelization framework for OpenAxiom

“…With the framework, we parallelized a software installation test, a set of OpenAxiom algebra library functions, and a user application on homotopy continuation method [2]. The experiments are run using a dual-core PC and a cluster.…”