Title :
BSPCloud: A Programming Model for Cloud Computing
Author :
Liu, Xiaodong ; Tong, Weiqin ; Hou, Yan
Author_Institution :
Sch. of Comput. Eng. & Sci., Shanghai Univ., Shanghai, China
Abstract :
Programming model for cloud computing has been a research focus recently. Some progresses have been made in cloud computing programming model, but little has been done on the model of predictable performance. In this paper, we introduce a programming model - BSPCloud, which has the advantages of performance predictability. A proof-of-concept of BSPCloud library has been implemented in java. We detail applying BSPCloud library on matrix multiplication, the experiment results display good speedup and scalability. We also analysis the factors which affect BSPCloud performance and the analysis also helpful to general parallel model of multi-core cluster.
Keywords :
Java; cloud computing; matrix multiplication; software performance evaluation; BSPCloud library; Java; cloud computing programming model; matrix multiplication; multicore cluster; performance predictability; Adaptation models; Cloud computing; Computational modeling; Message systems; Predictive models; Programming; Synchronization; BSPCloud; bulk synchronous parallel; cloud computing; programming model;
Conference_Titel :
Computer and Information Technology (CIT), 2012 IEEE 12th International Conference on
Conference_Location :
Chengdu
Print_ISBN :
978-1-4673-4873-7
DOI :
10.1109/CIT.2012.225
