WorkshopsHPC Asia 2020 will hold two workshops:
- IXPUG Workshop HPC Asia 2020
- Multi-scale, Multi-physics and Coupled Problems on Highly Parallel Systems (MMCP)
|Time||13:00 - 16:30, 17, January, 2020|
|Place||same venue of the conference|
Taisuke Boku (University of Tsukuba)
The Intel eXtreme Performance Users Group (IXPUG) is an active community led forum for sharing industry best practices, techniques, tools, etc. for maximizing efficiency on Intel platforms and products. IXPUG Workshop Asia 2020 is an open workship on high performance computing application, system and architectural with Intel technologies. This is a half-day workshop with invited talks and contibuted papers. The workshop aims to bring together software developers and technology experts to share challenges, experences and best-practice methods for the optimzation of HPC, Machine Learning and Data Analytics workloads on Intel Xeon Scalable Processors, Intel Xeon Phi Processors, Intel FPGA, and any related hardware/software platforms. The workshop will cover application performance and scalability challenges at all levels - from intra-node performace up to large-scale compute system. Any research aspect related to Intel HPC products is welcome to present in this workshop.
Workshop URL: https://www.ixpug.org/events/ixpug-hpc-asia-2020
Neda Ebrahimi Pour, M.Sc. (University of Siegen)
Prof. Dr-Ing. Sabine Roller (University of Siegen)
Prof. Dr. Nahid Emad (University de Versailles)
This workshop will provide a platform to present and discuss advances in numerical simulation for complex multi-scale, multi-physics and coupled problems. The goal is to gather researchers (computer scientists, engineers, mathematicians, physicists, chemists, biologists, material sciences etc.) specialized in different disciplines but related to challenges in multi-scale and multi-physics as well as coupled simulations on HPC systems.
The basis is - of course - the performance and efficiency of each component (region or solver), but the characteristics of other scales, other physical phenomena might change the game. Also a coupling tool might introduce e.g. load imbalances or hinder the overlapping of communication and computation. Applications with different characteristics in parts of the computational domain lead to additional performance issues. The optimum setting for the one part might be contradictory to the optimum for another; the overall optimum might be a non-optimal, but still good enough compromise for all individuals.
Several approaches have been considered, while developing different solutions depending on the application and hardware combination. On the application side monolithic or partitioned approaches have been introduced, on the hardware side homogeneous and heterogeneous cluster settings. All combinations have some advantages, some disadvantages, all leading to the question: how to find the optimal configuration and setting of all parameters, with respect to quality of solution vs. computational efficiency.
The basis is - of course - the performance and efficiency of each component (region or solver), but the characteristics of other scales, other physical phenomena might change the game. Also a coupling tool might introduce e.g. load imbalances or hinder the overlapping of communication and computation.
The main focus will be set on computational issues regarding performance and suitability for high-performance computing. Furthermore the underlying strategies to enable these simulations will be highlighted.
Keeping these aims in mind, contributions from all aspects of engineering applications will be considered. Topics of applications will include (but not be limited to):
- Multi-scale problems
- Multi-physics problems
- Molecular dynamics
- Multi-scale and/or multi-physics modelling for biomedical or biological systems
- Novel approaches to combine different scales and physics models in one problem solution
- Challenging applications in industry and academia, e.g. multi-phase flows, fllid-structure interactions, chemical engineering, material science, biophysics, automotive industry, …
- Load balancing
- Heterogeneous architectures
- New algorithms for parallel-distributed computing, specific to this topic.