 | Conference for Young Scientists Facing the Multicore-Challenge III September 19 - 21, 2012 Hochschule für Technik Stuttgart University of Applied Sciences with kind support by the German Gauß-Allianz e.V. |  | | |
Conference theme
The
prevalence of multicore and manycore technologies has brought ubiquitous
parallelism and a huge theoretical potential for compute-intensive
tasks. In theory, advancements in technology bring us closer to the
solution of the Grand
Challenges in
modern computing. In practice however, it is hard to achieve maximal
throughput in the results and to exploit all available capabilities.
Due to the inevitable paradigm shift towards multicore technologies,
parallelism is now affecting all kinds of software development
processes – from large-scale numerical simulation to desktop
commodity applications.
The trend for increased compute capability and more diverse hardware with its evermore complex topologies continues: In recent systems parallelism spreads over several levels of the system and heterogeneity is growing on the node as well as on the chip level. Data is kept across a nested hierarchy of memory stages and data locality becomes even more important. Most applications and algorithms are not yet ready to utilize available capabilities and a tremendous effort is required to close the gap. Different technologies and processing models, non-adjusted interfaces, and incomplete tool chains complicate holistic programming approaches and impede programmer’s productivity. On the other hand, resource contention, data conflicts and hardware bottlenecks keep performance away from theoretical peak. At the current state of the art in technologies and methodologies an interdisciplinary approach is required to tackle the obstacles in manycore computing. Only a comprehensive approach with contributions from computer science, applied mathematics, high performance computing, and engineering disciplines can face the multicore challenge. Compute- and memory-intensive applications can only benefit from the full hardware potential if all features on all system levels are taken into account in a holistic approach.
This conference is the 3rd in a series, starting in 2010 in Heidelberg at the Heidelberg Academy of Sciences (HAW) and 2011 at the Karlsruhe Institute of Technology (KIT) and the Engineering Mathematics and Computing Lab (EMCL). It aims to combine new aspects of multi-/manycore microprocessor technologies, parallel applications, numerical simulation, software development and tools. Contributions are welcome from all participating disciplines. Particular emphasis is placed on the support and advancement of young scientists.
The trend for increased compute capability and more diverse hardware with its evermore complex topologies continues: In recent systems parallelism spreads over several levels of the system and heterogeneity is growing on the node as well as on the chip level. Data is kept across a nested hierarchy of memory stages and data locality becomes even more important. Most applications and algorithms are not yet ready to utilize available capabilities and a tremendous effort is required to close the gap. Different technologies and processing models, non-adjusted interfaces, and incomplete tool chains complicate holistic programming approaches and impede programmer’s productivity. On the other hand, resource contention, data conflicts and hardware bottlenecks keep performance away from theoretical peak. At the current state of the art in technologies and methodologies an interdisciplinary approach is required to tackle the obstacles in manycore computing. Only a comprehensive approach with contributions from computer science, applied mathematics, high performance computing, and engineering disciplines can face the multicore challenge. Compute- and memory-intensive applications can only benefit from the full hardware potential if all features on all system levels are taken into account in a holistic approach.
This conference is the 3rd in a series, starting in 2010 in Heidelberg at the Heidelberg Academy of Sciences (HAW) and 2011 at the Karlsruhe Institute of Technology (KIT) and the Engineering Mathematics and Computing Lab (EMCL). It aims to combine new aspects of multi-/manycore microprocessor technologies, parallel applications, numerical simulation, software development and tools. Contributions are welcome from all participating disciplines. Particular emphasis is placed on the support and advancement of young scientists.
Submission topics
Topics of interest for conference submissions include (but are not limited to):
Emerging hardware architectures (Manycore, multicore-CPUs, GPGPUs, APUs, FPGAs, Cell, tiled manycore architectures, accelerators, ...)
- Aspects of microprocessor technologies and prospect of manycore technologies
- Parallel programming models, environments and languages (MPI, OpenMP, CUDA, OpenCL, PGAS, OpenACC, ...) for multicore computing
Hardware-aware computing and architecture-aware approaches for parallel numerical simulations, implementation and algorithm design
Heterogeneous computing, adaptive and reconfigurable computing
- Mathematical modeling, design of parallel algorithms and scalability issues
- Parallelization strategies in hybrid and hierarchical setups
- Autotuning strategies; portability and flexibility of software solutions
Library and tool support for parallel applications
- Performance analysis and modeling
- Novel programming tools for parallel programming
Compiler techniques and code optimization strategies for parallel systems
- Virtualization strategies and hardware transparency
Memory behavior analysis, parallel data access and parallel data structures
Efficient parallel numerical methods
- Parallel applications (e.g. numerical simulation, image processing)
- Parallelism for nun-numerical and non-HPC applications
Practice and experience of multicore programming
Benchmarking results of scientific applications, multicore application studies
- Interdisciplinary approaches and theoretical concepts
Conference language is English.
Invited Speakers 2012
- Bernd Winkelsträter, Fujitsu: "On the Road to ExaFlops - Some Reality Checks"
- Richard Barrett, Sandia National Labs: "Exascale Preparations using Miniapps"
- Torsten Höfler, ETH Zürich: "MPI-3.0: A Response to New Challenges in Hardware and Software"
- Mondrian Nüssle, EXTOLL: "DEEP and EXTOLL, the scalable interconnect for the DEEP Booster architecture"
Tutorials on multicore technologies and parallel programming aspects will held by
- Stefan Andersson, Cray, "Large-scale Computing"
- Royd Lüdtke, Rogue Wave, "Rogue Wave ThreadSpotter - A Statistic Approach of Cache Related Performance Analysis"
- Michael Klemm, Intel, "The Intel Xeon Phi Coprocessor Architecture and Software Environment"
Short Talks and Poster Session
The conference program is complemented by poster sessions and short talks of young scientists on ongoing research.
Awards and Travel Support
We are happy to announce Best-Paper-Awards, a Best-Poster-Award, a Best-Presentation-Award and Travel Support for young scientists kindly granted by the German Gauß-Allianz. See this page for further details.
Third Conference: Facing the Multicore-Challenge III
After fruitful conferences at the Heidelberg Academy of Sciences in 2010, at Karlsruhe Institute of Technology (KIT) in 2011, and the multitude of open topics, the third conference, Facing the Multicore-Challenge III, will take place at the Hochschule für Technik, Stuttgart, Germany, September 19-21, 2012.
Questions or Comments
In case of any questions, please contact the organizers by sending your request to info@multicore-challenge.org.