University gives Java parallelism a boost

Researchers at the University of Illinois at Urbana-Champaign have released DPJizer, a tool geared for parallel programming

Java developers are getting more assistance in writing parallel programs for multi-core systems with another technology from the University of Illinois at Urbana-Champaign.

The university said this week that computer science researchers at the school have released an interactive tool, called DPJizer, to simplify writing safe parallel programs in DPJ (Deterministic Parallell Java), a Java-based modern type and effect system detailed by the university earlier this fall.

[ A planned upgrade to the official Java specification features multicore capabilities. | Keep up with the latest Java developments with the JavaWorld Enterprise Java newsletter. ]

DPJizer is offered as an Eclipse plug-in and is described as the first interactive, practical type and effect inference tool for a modern object-oriented effect system. An effect system, the university said, requires extensive annotations from programmers who should strive to keep annotations in sync with code. DPJizer saves time by performing a  whole-program analysis to infer some DPJ annotations automatically, with many being more precise than those written manually. Also, programmers can review and change annotations inferred by DPJizer.

"DPJizer increases the productivity of programmers in writing safe and deterministic-by-default parallel programs for multi-core systems," said Mohsen Vakilian of the Illinois research team. The tool makes it much easier to write and maintain DPJ programs, Vakilian said.

Instructions for downloading DPJizer can be found at the DPJizer wiki.

The DPJ project is intended to provide deterministic-by-default semantics for an object-oriented, imperative parallel language using primarily compile-time checking. Deterministic means the program produces the same visible output for a given input in all executions. By-default means that deterministic behavior is guaranteed unless a programmer requests non-determinism. By contrast, shared-memory programming models available today are inherently non-deterministic and can have undetected data races. DPJ has featured a compiler, runtime, and related materials as open source software.

This article, "University gives Java parallelism a boost," was originally published at InfoWorld.com. Follow the latest developments in business technology news and get a digest of the key stories each day in the InfoWorld Daily newsletter.

From CIO: 8 Free Online Courses to Grow Your Tech Skills
Join the discussion
Be the first to comment on this article. Our Commenting Policies