Many data-intensive scientific applications are irregular by nature. They may present irregular data structures, control flow or communication. Irregular applications pertain both to well-established and emerging fields, such as Computer-Aided Design (CAD), bioinformatics, semantic graph databases, social network analysis and security. These applications typically use pointer-based data structures such as graphs and trees, often present fine-grained synchronisation and communication, and generally operate on very large data sets.
Current supercomputing systems are organised around components optimised for data locality and regular computation. Developing irregular applications on them demands a substantial effort, and often leads to poor scaling and performance due to their limits with fine-grained synchronisation and small data transfers. However, executing these applications efficiently will be a key requirement for future systems.
The solutions needed to address their challenges can only be found by considering the problem from all the points of view, from micro- to system-architectures, from compilers to languages, from libraries to runtimes, up to rethinking how algorithms operate.
Only collaborative efforts among researchers with different profiles, including end users, domain experts and computer scientists, can lead to significant breakthroughs.
This special issue aims at identifying, outlining and systematising the set of challenges and the state-of-the-art approaches for irregular applications that today appear in many different venues at many different levels (e.g. algorithms, architectures, compilers, languages, libraries), but really are strongly connected by the theme of irregularity.
The issue will carry revised and substantially extended versions of selected papers presented at the SC'11 Workshop on Irregular Applications: Architectures & Algorithms (IA3) and at the HiPC Workshop on Parallel Algorithms and Software for Analysis of Massive Graphs (ParGraph), but we also strongly encourage researchers who did not participate in these events to submit papers for this call.
Suitable topics, of both theoretical and practical significance for irregular applications, include but are not limited to:
- Micro- and system-architectures
- Network and memory architectures
- Heterogeneous and custom architectures (GPUs, FPGAs)
- Modelling and evaluation of architectures
- Innovative algorithms
- Parallelisation techniques and data structures
- Languages and programming models
- Library and runtime support
- Compiler and analysis techniques
- Combinatorial (graph) algorithms and their applications
Submission due: 15 March, 2012
First review results: 30 April, 2012
Final copy deadline: 25 June, 2012