Digital Agenda for Europe
A Europe 2020 Initiative

Software tools for next generation computing

Article
The objective of this workshop is to discuss the state of the art and the future perspectives in the area of software tools and methods, in view of the expected evolution of computing hardware. Discussion will be focused on software for cyber-physical systems, industrial and time-critical applications, with typical application areas in aerospace, automotive, transport and automation sectors.A small number of experts will be asked to advice the European Commission on future actions and directions for the research and innovation work programme.
Share this
Date: 
24/06/2014
Venue: 
Universitaire Stichting - Fondation Universitaire, Egmontstraat / Rue d'Egmont 11 - 1000 Brussel / Bruxelles
Organiser: 
HiPEAC with the support of the European Commission

Computing market and applications evolve quickly over time, and some trends are clearly visible in the application areas of CPS, industrial and time-critical systems.

  • The computing market is progressing towards very heterogeneous platforms, where different types of computing chips (CPUs, GPUs, DSPs, FPGAs) may work together in the same system, and a specific computing application has to be optimised against orthogonal or conflicting criteria (energy consumption, throughput, response time, guaranteed time behaviour, memory footprint, processing load, fault tolerance, resiliency…) Current tools and languages do not handle very efficiently those constraints.

 

  • Parallel programming is more and more needed in many application areas, but the most widely used programming languages do not handle natively parallelism, while specialised frameworks and languages designed for harnessing the power of massively parallel hardware have still a limited impact on the market.

 

  • The level of abstraction offered by current tools is in most cases not good enough: in order to fully exploit the possibilities offered by recent hardware, developers should have deep knowledge of both the application domain and last generation computing architectures, and master many different tools which have steep learning curves. The complexity of complete applications is exploding and reaches the limit of understanding by developers. Can this problem be mitigated by new levels of abstraction?

 

  • Some European projects have produced important results in this area, but the overall impact at portfolio level is not satisfactory.

 

The workshop  will be held on June 24th, 2014- from 10:00 to 17:00. It is organised by HiPEAC  with the support of the European Commission (Complex Systems and Advanced Computing). It is "by invitation only" workshop.

Last updated on 19/07/2015