EPSRC logo

Details of Grant 

EPSRC Reference: GR/J82423/01
Title: EXTENSION TO UC++ TO SUPPORT FINE GRAIN AND SIMD PARALLELISM
Principal Investigator: Winder, Dr R
Other Investigators:
Dzwig, Dr P
Researcher Co-Investigators:
Project Partners:
Department: Computer Science
Organisation: UCL
Scheme: Standard Research (Pre-FEC)
Starts: 01 September 1994 Ends: 30 November 1996 Value (£): 211,679
EPSRC Research Topic Classifications:
Parallel Computing
EPSRC Industrial Sector Classifications:
Related Grants:
Panel History:  
Summary on Grant Application Form
UC++ is an minimalistic extension of C++ to support active object parallelism (i.e. MIMD parallelism) using an object allocation rather than an object type strategy. The current implementation of UC++provides course-grain parallelism capability. The aim of this project is to investigate the inclusion within UC++ of features to support data parallelism (for both SIMD and MIMD machines) and possibly also fine-grain (expression-level) parallelism: The intention being to support a common programmer interface for all varieties of parallelism; SIMD and MIMD parallel machines and networks of machines. The major objectives of the project are to provide a library of types which enables data parallel computation in UC++ without further extension of the base language and to show that these permit the easy implementation of data parallel systems integrated with object-level parallel systems using the same language. Progress:Due to a minor difficulty over staff start date, the project started fully on 1st October 1994 and has therefore been running for 5 months. A full survey of the principal C and C++ based data parallel languages has been completed and a survey paper is in preparation which will be submitted for publication to one of the leading journals in the field.Fortran 90 and HPF have also been investigated to ascertain useful properties for the data structures used in data parallel computing. Although the survey has focussed on data parallelism, material on all extensions to C and C++ aimed at supporting parallelism of all varieties has been collected.The UC++ compiler as delivered to this project was not entirely complete so some work has been required on various tidying up aspects on the parser and on completing ports of the compiler to the machines that provide the infrastructure base for this project. This aspect of the work is now nearly complete. An initial skeleton of the design of the types and library are now being put in place. Time has also been spent in coordinating with the Europa working group. This is a CEC funded working group, run by LPAC, investigating the possibility of defining a de facto European standard for a parallel C++, supported by the major European manufacturers.
Key Findings
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Potential use in non-academic contexts
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Impacts
Description This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Summary
Date Materialised
Sectors submitted by the Researcher
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
Project URL:  
Further Information:  
Organisation Website: