We present solutions to statically load-balance scatter operations in parallel codes run on Grids. Our loadbalancing strategy is based on the modification of the data distributions used in scatter operations. We need to modify the user source code, but we want to keep the code as close as possible to the original. Hence, we study the replacement of scatter operations with a parameterized scatter, allowing a custom distribution of data. The paper presents: 1) a general algorithm which finds an optimal distribution of data across processors; 2) a quicker guaranteed heuristic relying on hypotheses on communications and computations; 3) a policy on the ordering of the processors. Experimental results with an MPI scientific code of seismic tomography illustrate the benefits obtained from our load-balancing.

Scientific discovery can be accelerated through computation and visualization. This acceleration results from the synergy of expertise, computing tools and hardware for enabling high-performance computation, information science and visualization that is provided by a team of computation and visualization scientists collaborating in a peer-to-peer effort with the research scientists.

We present load-balancing strategies to improve the performances of parallel MPI applications running in a Grid environment. We analyze the data distribution constraints found in two scientific codes and propose adapted code transformations to load-balance computations. Experimental results confirm that such source code transformations can improve Grid application performances.

We introduce a parallelization of the maximumlikelihood cosine transform. This transform consists of a computationally intensive iterative fitting process, but is readily decomposed for parallel processing. The parallel implementation is not only scalable, but has also brought the execution time of this previously intractable problem to feasible levels using contemporary and cost-efficient highperformance computers, including an SGI Origin 2000, an SGI Onyx, and a cluster of Intel-based PCs. Key words : parallel processing, emission spectroscopy, cosine transform, maximum-likelihood inversion, performance evaluation, DParLib, MPI. 1 Introduction Michelson interferometers used in optical spectroscopy produce interferograms that can be considered a superposition of cosine functions. The optical spectrum is estimated from this interferogram using some transform method, typically the Fourier transform, a least-squares procedure, which presumes normally-distributed noise. A more general ...

We present load-balancing strategies to improve performances of parallel MPI applications running in a Grid environment. We analyze the data distribution constraints found in two scientic codes and propose adapted code transformations to load-balance computations. We present the general framework for our load-balancing techniques. We then describe these techniques as well as their application to our examples. Experimental results conrm that adapted source code transformations can improve Grid application performances.