Paul Kelly: Current/recent research grants
Current research grants
- IBM Faculty Award "Toward Autonomic Componentized
Software Overlays"
- Field-Programmable Logic for Custom Computing (platform
grant)
5 years from 2005. With Wayne Luk (PI), Peter Cheung, George
Constantinides, Oskar Mencer.
- HiPEAC: High Performance Embedded Architectures and
Compilers (Network of Excellence)
5 years from 2004. (member of large consortium, funding distributed
over project lifetime).
- Static and Dynamic Program Optimisation
British Council ARC travel grant, to collaborate with Prof Christian
Lengauer at the University of Passau, Germany. 2 years from 2004.
Past funding
- Framework for domain-specific optimization at
run-time
Microsoft "Rotor" award, 1 year from 2004.
- Extensible domain-specific performance analysis using
dynamic aspect weaving
IBM Eclipse Innovation Grant, 1 year from 2004.
- DESORMI:
Delayed-evaluation, self-optimising Java Remote Method Invocation.
1 RA for 3 years (2001 onwards), EPSRC.
(jointly with Tony Field)
See project page: http://www.doc.ic.ac.uk/~samt/DESORMI/index.html
- OSCAR:
Optimising scientific applications at run-time.
1 RA for 3 years (2001 onwards), EPSRC.
(jointly with Tony Field)
- Compaqt:
Combining Program and Query Transformation for the Efficient Exploitation of
Parallel Database Hardware.
1 RA for 3 years (1993 onwards), EPSRC.
(Jointly with Tony Field and Hessam Koshnevisan, Imperial).
- Cramp:
Combining randomisation and mixed-policy caching for bounded-contention shared-memory.
1 RA for 3 years (1994 onwards), EPSRC.
(Jointly with Tony Field, Imperial).
Objectives:
- To study the design of general-purpose shared-memory parallel
computers.
- to design data management protocols and policies with high average
performance and reasonable worst-case performance,
- to evaluate them using benchmark programs, and
- to establish analytical bounds on the performance characteristics
of different designs.
Click here for
further details.
- Futurespace
The Optimisation of Coherency and Communication in Parallel Systems using Shared Abstract
Data Types.
1 RA for 3 years (1994 onwards), EPSRC.
(Jointly with Tony Field, Imperial).
Objectives:
- Extend the concept of shared memory to abstract data types
- Design and implement a selection of shared ADT library functions
- Evaluate alternative distribution, replication and scheduling
techniques using simulation and analytical modeling
- Demonstrate use of shared ADTs in example applications codes
running under simulation and on available parallel equipment.