The HPC Connection Workshop is an international High Performance Computing summit organized by the Asia Supercomputer Community and Inspur Group. This event takes place three times a year: during the ASC in China, the ISC in Germany, and the SC in America. We are committed to the exploration of application innovation and current trends in emerging fields of science and engineering. During the workshop, top researchers and leading professionals from around the world gather together to discuss disruptive technologies and latest development in supercomputing.
Previous workshops http://www.asc-events.org/ASC16/Workshop.php http://www.asc-events.org/index.html
Director, Leibniz Supercomputing Centre
Professor, Chair of Computer Engineering and Computer Organization, Technical University of Munich
|The Energy Challenge in HPC
To solve the energy challenge of future Exascale systems, a holistic approach including data center infrastructure, HPC system architecture and software as well as application algorithms is needed. With SuperMUC in the context of the Leibniz Supercomputer Center and the GCS infrastructure, first steps in this direction have been taken. The presentation describes the techniques including direct warm water cooling, reuse of waste energy, intelligent middleware using frequency scaling, tools for monitoring, analysis, optimization and prediction as well as high level application support and extreme scaling workshops.
Professor, Director of High Performance Computing Education, Iowa State University
|Evaluation of Allinea's and Intel's Performance Tools for HPC Code Optimization
HPC is becoming increasingly important for not only the commercial and industrial world but also for pure and applied research. Training developers how to optimize applications to provide high performance is critical for those who use HPC machines. To be able to optimize applications, it is necessary to have performance tools that are easy-to-use and provide the needed functionality. The purpose of this work is to evaluate and compare Allinea's and Intel's performance tools to aid in application optimization for use in HPC.
Professor, Karlsruhe Institute of Technology
|Algorithm Engineering for Large Data Sets
The talk will introduce the methodology of algorithm engineering and explains why this this methodology is important for grand challenge applications of computers. I will then proceed with examples from my own work including sorting, graph partitioning and other selected topics relevant for massively parallel computing.
Director of the Architecture Design and Embedded Software Division, French Alternative Energies and Atomic Energy Commission
|New Software and Hardware Architecture Solutions for HPC and BigData
The exponential growth of information required for fine grain simulations and visualization or generated by humans and objects (BigData), creates an explosion of digital data and the need for fast data processing and analysis while keeping a constant energy envelope and cost of ownership. This will cause a paradigm shift for HPC and BigData infrastructures to move from a tight focus on performance towards energy-efficiency and total cost of ownership. The components of future HPC and BigData infrastructures and their integration into a full system should be reconsidered. This will be made possible with the emergence of new domain specific highly parallelizable software and hardware, and more generic solutions to increase the compute density and energy efficiency, all these solutions minimizing the cost of data movement and optimizing the development cost of new Integrated Circuit design and manufacturing.
Professor, University of Paderborn
Director of Computational Materials Design, Max Planck Institute for Iron Research
|HPC-Based Quantum Mechanical Design of Engineering Materials
Modern engineering materials have evolved from simple single phase materials to nano-composites that employ dynamic mechanisms down to the atomistic scale. HPC-based quantum mechanical approaches provide perfect tools to new design routes but face serious challenges. Specifically, to obtain accurate predictions an extreme high accuracy in the thermodynamic (statistical) averages is critical and requires both new methodological approaches and an optimization towards modern HPC architectures. The predictive power of these approaches and the impact they can have on developing new strategies in materials design will be discussed for examples ranging from modern ultra-high strength steels to light weight metallic alloys.
Professor, Beihang University, Sun Yat-sen University
Director of the Key Project on HPC, National High-Tech R&D Program
|China HPC Development in the 13th 5-Year Plan
After a brief review on HPC research and development in China’s high-tech R&D program, this talk will introduce the plan of HPC development under the new Key R&D Program of China in the 13th 5-year plan. The major challenges and opportunities in establishing the eco-system for high performance computing in China will be discussed, including the technical issues in developing the next generation HPC system using domestic processors, the urgency of developing application software for the domestic processor-based system, and the difficulty in establishing a national HPC environment with sustainable operation model and mechanism. The goal and the major activities of the new key project on HPC will be briefly presented.
National Supercomputing Center in Wuxi
|The New Sunway Supercomputer: System & Applications -- Introduction to the National Supercomputing Center in Wuxi and Its Collaboration with Tsinghua University
Supported by China's National High Technology R&D Program in the 12th five-year plan, in this summer, we would see the opening of a new supercomputer system at the new center in Wuxi. As joint center between the City of Wuxi, Jiangsu Province, and Tsinghua University, the National Supercomputing Center in Wuxi would be run by a team from Tsinghua. In this talk, we would provide a brief introduction to the supercomputer system, as well as the preliminary results that we have achieved in key application domains, such as earth system modeling, seismic inversion, and deep learning.
Institute of Software, Chinese Academy of Sciences
|The New Sunway Supercomputer: System & Applications -- Early Experience on Optimizations of Application Codes on the New Sunway Supercomputer
This talk will share some of our early experiences of using the new Sunway Supercomputer, including the development of a high-performance extended math library xMATH, the optimization of the emerging benchmark HPCG and the deployment of an implicit nonhydrostatic atmospheric dynamic core. Some key techniques that we use to achieve high performance on the new architecture at the scale of millions of hybrid cores will be presented.
Computer Network Information Center, Chinese Academy of Sciences
|The New Sunway Supercomputer: System & Applications -- Large-scale Phase Field Simulation for Coarsening Dynamics Based on Cahn-Hilliard Equation with Degenerated Mobility
The highly nonlinear and severely stiff Cahn-Hilliard equations with degenerated mobility for microstructure evolution are solved at extreme scale, demonstrating that the latest advent of high performance computing platform and the new advances in algorithm design are now offering us the possibility to accurately simulate the coarsening dynamics at unprecedented spatial and time scales. The code has demonstrated good strong and weak scalability and has achieved 39.7 PFlops in double precision for the largest configuration, using over 9 million cores.
Institute of Process Engineering, Chinese Academy of Sciences
|The New Sunway Supercomputer: System & Applications -- Large-scale Molecular Dynamics Simulation of Thermal Conductivity of Silicon Nanowires on the New Sunway Supercomputer
Based on a powerful and efficient nonequilibrium molecular dynamics (NEMD) simulation method, the computation of thermal conductivity of silicon nanowires is realized to overcome the known finite size effects, where the longitudinal dimensions of the simulations significantly exceed the micrometer scale. More importantly, the lateral characteristic sizes are much larger than 10 nanometers, explicitly comparable with the silicon nanowires fabricated and measured experimentally, whereas the traditional simulation size is of several nanometers. The powerful virtual experimental measurement implemented on the Sunway supercomputer can achieve the direct prediction of thermal conductivity of real-scale silicon nanowires, and delineate the longitudinal size dependence of their thermal conductivities.
Professor, George Washington University
Director, GWU High Performance Computing Lab
Founding Director, GWU Institute for Massively Parallel Applications and Computing Technologies
Founding Co-Director, NSF Center for High-Performance Reconfigurable Computing
|Exploiting Hierarchical Locality for Extreme Scale Computing
Modern high-performance computer architectures are characterized with massive hardware parallelism and deep hierarchies. Extreme scale and exascale systems will aggravate this problem. Locality exploitation at all levels of the hierarchy is a must as the relative cost of data transfer is increasing in latency and power. Locality-aware programming paradigms can provide programmers with the tools to exploit data locality. However, asking programmers to worry about expressing locality relations at multiple architecture hierarchy levels in their program is detrimental to productivity. In this talk I address how to provide productive support for exploiting hierarchical locality through run-time systems and hardware.
* By invitation only.