A preview of the PDF is not available
PCJ - a PGAS library for Java
This paper presents the performance of the PCJ -- a new library for Java language that helps to perform parallel and distributed calculations. The library is motivated by partitioned global address space (PGAS) model represented by Co-Array Fortran, Unified Parallel C or Titanium. PCJ has ability to work on the multicore systems hiding details of inter- and intranode communication. In the PCJ, each task has its own local memory and stores and access variables only locally. Some variables can be shared between tasks and that variables can be accessed, read and modified by other tasks. The library provides methods to perform basic operations like synchronization of tasks, get and put values in asynchronous one-sided way. Additionally the library offers methods for creating groups of tasks, broadcasting and monitoring variables. The PCJ library fully complies with Java standards therefore the programmer does not have to use additional libraries, which are not part of the standard Java distribution. In this paper the PCJ library with the support for multicore, multithreaded systems is presented. It has been used to run example HPC benchmark suite applications on the multicore nodes. The results show good performance and scalability compared to native MPI and OpenMP implementations of the same algorithms. It is noteworthy that the PCJ library has great promise to be successful in scientific applications.