作者: Andrew Chien , Brad Calder , Stephen Elbert , Karan Bhatia
DOI: 10.1016/S0743-7315(03)00006-6
关键词:
摘要: The exploitation of idle cycles on pervasive desktop PC systems offers the opportunity to increase available computing power by orders magnitude (10×-1000×). However, for distributed be widely accepted within enterprise, must achieve high levels efficiency, robustness, security, scalability, manageability, unobtrusiveness, and openness/ ease application integration.We describe Entropia system as a case study, detailing its internal architecture philosophy in attacking these key problems. Key aspects include use of: (1) binary sandboxing technology security (2) layered scalability (3) an open integration model allow applications from many sources incorporated.Typical System includes molecular docking, sequence analysis, chemical structure modeling, ;and risk management. come diverse set domains including virtual screening drug discovery, genomics targeting, material property prediction, portfolio In all cases, scale thousands nodes have no dependences between tasks. We present representative performance results several that illustrate performance, linear scaling, overall capability presented system.