Using MPI: portable parallel programming with the message-passing interface
Using MPI: portable parallel programming with the message-passing interface
Enabling Java for high-performance computing
Communications of the ACM
Java Native Interface: Programmer's Guide and Reference
Java Native Interface: Programmer's Guide and Reference
The Java Language Specification
The Java Language Specification
Computer Languages, Systems and Structures
Reducing memory sharing overheads in distributed JVMs
HPCC'05 Proceedings of the First international conference on High Performance Computing and Communications
Hi-index | 0.00 |
Java has emerged as a possible lingua franca for uniting the disparate worlds of web computing, cluster computing, multiprocessor computing, and uniprocessor computing. In this paper we explore Java's support for cluster computing. We compare the two basic approaches to utilizing Java in a cluster. The first approach views the cluster as a single computer running a single Java virtual machine. The second approach connects the standard Message Passing Interface to Java and therefore views the cluster as a collection of computers each running a separate Java virtual machine. We evaluate the performance of the two approaches on a set of benchmark applications obtained from the Java Grande Forum.