Abstract

Traditionally, scientists have been using individual workstations, in-house computational clusters, or campus high-performance resources to conduct their scientific computations. When necessary, they applied for cycles on the top national resources hosted by supercomputing centers and national laboratories. In the past two decades, with the advancement of broad resource sharing technologies such as Condor [1] and Globus [2], efforts such as SETI@ home [3], and more recently Cloud technologies [4], scientists are faced with an ever-expanding choice of computing platforms each with its own benefits and drawbacks related to performance, cost, ease of use, and other characteristics. Having this significant number of computational options is a great opportunity for computational sciences providing resources that can scale up with the ever-expanding data collections and everincreasing computational needs of …

Date

July 19, 2010

Authors

Raphael Bolze, Ewa Deelman

Journal

Cloud Computing and Software Services

Pages

173

Publisher

CRC Press