Software sytems used in domains such as automotive, space, large scientific infrastructures or smart cities, have seen their complexity drastically increased in the last years due to the need of implementing advanced functionalities applied to cutting edge research areas like autonomous driving, traffic management, etc. The performance requirements of these new functionalities urge the use of parallel and heterogeneous architectures capable of delivering the expected throughput while still considering other non-functional aspects of the system like real-time requirements, fault-tolerance, and energy budgets, among others. The research activities will be focused on one or several of the following topics, among others:1. In close collaboration with leading automotive and space companies, development of software system prototypes to evaluate the new developments
2. Optimization of advanced software functionalities targeting parallel computing architectures, including multi-core systems, GPUs and/or FPGAs.
3. Programming models: use, implementation and extension of programming models with parallel capabilities, like OpenMP and CUDA, to target complex software systems.
4. Code transformation and optimization: high-level synthesis and compilation techniques to boost performance, predictability and correctness, mostly based on LLVM.
5. Runtime systems: run-time techniques to boost performance, heterogeneity (targeting GPUs and FPGAs) and fault-tolerance.
Key Duties
Research on parallel programming models towards productivity and other non-functional requirements
Support the development of compiler and run-time techniques to support extensions for performance and other non-functional requirements.
Support the Development of advanced application prototypes for the automotive and space domains to evaluate the proposed techniques
Write technical reports and papers
Attend technical meetings and conferences
Termini per aplicar: Dissabte, 31 Desembre, 2022
Més informació
Més entrades de Beques
