Keywords: Concurrent and Distributed Software, Object Oriented Software, Software Engineering, High Performance Networks and Libraries, Middleware, Runtime Environments, Patterns.
In the last years there have been constant improvements of computing hardware: there are now multi-core CPUs (with 8-16 cores), the introduction of hardware accelerators, GPUs and the upcoming Intel MIC, for general purpose computing relying on their inherently parallel nature, with dozens and hundreds of cores, and finally, cloud computing can provision a virtually infinite number of compute nodes. The software to solve a given (scientific) problem must be adaptable to this hardware heterogeneity and find new ways to make efficient use of such distributed compute resources. Additionally, how to program these systems is constantly open to discussion, as the options range from traditional approaches, based in MPI for distributed memory processing and OpenMP for shared memory, to the present research and development in object oriented approaches, which can provide multiple software engineering advantages in this field and are the focus of the workshop.
The EOOPS workshop provides a forum for engineers, researchers, and scientists to exchange the latest results on object oriented programming methodologies to match the challenges posed by the developing hardware and application requirements.
Topics which are interesting for the workshop include, but are not limited to:
Object oriented software in HPC and scientific programming
Support of object oriented sofware for hardware accelerators
Novel approaches in object oriented communication software
Object oriented message-passing middleware
Grid and cloud computing software and systems
Emphasis on object oriented concurrent methodologies (not limited to FORTRAN and C/C++)
Object oriented solutions for grid and cloud computing
Aspect oriented software, Scala, Erlang, Akka
Interoperability of libraries, API design for parallel libraries
Performance evaluation and analysis
Unit tests and contineous integration tests for parallel and distributed applications
Patterns and frameworks
Case studies, examples of important designs
Papers must describe original work, and must not have been accepted or submitted for publication elsewhere. Accepted papers will be published in the conference proceedings, which will be published by IEEE Computer Society.
The EOOPS workshop started in 2013 at IEEE AINA conference in Barcelona. The accepted papers list and the program can be found here: past EOOPS.
Please, submit papers to other AINA workshops or the main AINA conference via the AINA-2014 website (http://www.aina-conference.org/2014/).
Heinz Kredel (IT Center, University of Mannheim, Germany)
Guillermo Lopez Taboada (University of Coruña, Spain)
Markus Aleksy (ABB AG, Germany)
to be announced