K. Vikram, K. Avijit, and S.K. Aggarwal (India)
: Grid Computing, Parallelizing Compilers, Dependence Analysis
Grid Computing aims at providing computing resources much in the same way as electricity is provided in a power grid. A grid is a large collection of computing resources that are managed by a resource manager to make it appear like a single powerful computing resource. The technol ogy is still in its developing stages and has not yet ma tured to the point that a sequential task submitted to the grid can run on more than a single machine simultaneously. Automatic parallelization techniques have been proposed, that can parallelize sequential tasks into multiple threads that can be executed simultaneously with some communi cation between each other. Nevertheless, actual tools that apply these methods to computational grid, are still rare. In this paper, we discuss certain program parallelization techniques and propose a transformation to parallelize se quential programs for a grid. We discuss our tool, which is based on the SUIF. Our tool extends parallelization strat egy of SUIF to convert sequential code into its semantic equivalent parallel code that can run on a grid.
Important Links:
Go Back
