CARP: Correct and Efficient Accelerator Programming
Funded by:
European Commission FP7
Duration: December 2011
until December 2014
Summary of the project
The CARP European research project aims at improving the programmability of accelerated systems, particularly systems accelerated with GPUs, at all levels. This involve designing high-level programming formalisms geared towards accelerators, writing highly optimizing compilers to compile high-level code into efficient OpenCL, verifying correctness of accelerator kernels, and employing intensive symbolic testing techniques to find bugs. See the objectives page for more details.
Partners
The CARP project unites world-class European research institutions and innovative industrial partners:
- Imperial College London (coordinator), United Kingdom, Key personel: Alastair Donaldson, Multicore programming models, verification
- ARM, United Kingdom, Key personel: Anton Lokhmotov, GPU hardware, OpenCL tools
- Realeyes, Estonia, Key personel: Elnar Hajiyev, Eye-tracking software
- RWTH Aachen University, Germany, Key personel: Joost-Pieter Katoen, Quantitative analysis
- Monoidics, United Kingdom, Key personel: Dino Distefano, Program analysis
- Twente University, Netherlands, Key personel: Marieke Huisman, Formal verification
- ENS, France, Key personel: Albert Cohen, Polyhedral compilation
- Rightware, Finland, Key personel: Teemu Uotila, GPU benchmarking