The 2011 International Conference on High Performance Computing & Simulation
Tuesday Keynote I: High-Performance Computing and Biology: The Quest for A Predictive Biological Theory
Rick L. Stevens
Argonne National Laboratory and The University of Chicago, Illinois, USA
In this talk we will discuss the role that high-performance computing and advanced data systems play in accelerating the transition of biology from a science primarily focused on description and explanation to a new science focused on systems level understanding and data driven predictive theories. We will discuss the role that genomics driven biological models are playing in reshaping what is possible in biological analysis and how high-performance computing technologies are playing a critical role in enabling new types of biological investigations. We will outline how we go from raw DNA sequence data to predictive whole genome models of biological phenotype and how automated annotation and analysis systems are enabling small biological labs to do things that were not possible at any scale just a few years ago. We will project how the increases in computing capability coupled with the increases in biological sequence generation throughput and the rise of data integration methods will result in a rapid expansion of our understanding. If time permits, we will discuss the impact that predictive biological models can have on health, energy, the environment and science.
Tuesday Keynote II: Machines, Methods and Music: On the Evolution of e-Research
David De Roure
Oxford e-Research Centre, University of Oxford, United Kingdom
10 years ago we saw researchers in many disciplines beginning to adopt e-Science technologies; now, against a backdrop of increasing computational capability, we see acceleration of research through broader adoption and sharing of tools, techniques and resources, both for 'big science' and the 'long tail scientist'. This talk will discuss the evolution of e-Research, focusing on a web-scale computational musicology project as an illustration of emerging methodology and using the myExperiment social website as a lens to glimpse future research practice.
Wednesday Keynote I: Embracing Heterogeneity in High-Performance Computing
The George Washington University, Washington, D.C., USA
In recent years, the top supercomputers in the world were built around the use of heterogeneous architectures where conventional microprocessors are accelerated using specialized processors. Those specialized processors included gaming processors, general purpose graphics processors (GPGPUs) and field programmable gate arrays (FPGAs). Such heterogeneous high-performance computing (HPC) systems have achieved unprecedented successes. In June of 2008, the Road Runner, a system at Los Alamos National Labs, using the Cell processor which was originally built by IBM/Sony/Toshiba for the Sony PlayStation 3, was able to reach the PetaFLOPs performance range for the first time. The Road Runner system used 6,912 dual-core AMD Opteron® chips and 12,960 Cell chips. In November of 2010, China was able to build, also for the first time in history, a supercomputer, Tianhe-1A, that ranks as the top supercomputer in the world. Tianhe-1A major speed boost came from 7000 NVIDIA graphical processing units (GPUs). There are, therefore, obvious performance advantages for such heterogeneous architectures. However, there are also enormous productivity challenges and developing applications for such systems requires heroic programming efforts, often using proprietary tools. Portability is yet another major problem as such complex codes may not readily run on other parallel machines. In this talk, we systematically examine the productivity challenges for heterogeneous parallel computers. Research directions and potential solutions are identified along with promising early efforts aiming towards making heterogeneity a first-class citizen in the HPC realm. This is all examined in the light of the needs of domain scientists who wish to focus on their science and develop efficient applications without having to attain specialized architectural knowledge.
Wednesday Keynote II: Challenges in Hybrid and Federated Cloud Computing
Ignacio M. Llorente
OpenNebula Project Director, DSA-Research.org, Universidad Complutense de Madrid, Spain
The keynote describes the different cloud federation scenarios, ranging from a federation built on commercial cloud providers that offer no real support for federation to one built on data centers of the same organization where the sites are completely dedicated to supporting all aspects of federation. The level of federation is defined based on the amount of information disclosed and how much control over the resources is provided across sites. The talk also presents the existing challenges for interoperability in federated and hybrid cloud computing scenarios, and ends with real-life examples of multi-cloud environments running OpenNebula.
Thursday Keynote I: High Performance Computing Efforts of Turkey in Science and Engineering
National Center for High Performance Computing, Ulusal Yüksek Başarımlı Hesaplama Merkezi (UYBHM), Istanbul, Turkey
This talk will provide introduction to İstanbul Technical University and National Center for High Performance Computing (UYBHM), followed by brief information about research groups that are using high performance computing infrastructure in Turkey, specifically in UYBHM. Capabilities and opportunities that are currently offered to researchers along with UYBHM and Turkey’s efforts in Partnership for Advanced Computing in Europe (PRACE) will be summarized, and results of some sample research project will be presented.
Thursday Keynote II: Entering the Era of Petascale Computing in Europe
Institute for Advanced Simulation, Jülich Supercomputing Centre, and Fachbereich Mathematik und Naturwissenschaften, Bergische Universität Wuppertal, 42097 Wuppertal, Germany
Within the last two years, a consortium of 20 European countries is preparing the legal and technical prerequisites for the establishment of a leadership-class supercomputing infrastructure in Europe. The consortium named "Partnership for Advanced Computing in Europe" signed its statutes in April 2010. Four PRACE members have committed to provide compute cycles worth €100 Million each within a 5 years period until 2015 on machines of the highest performance class (Tier-0), providing a diversity of architectures beyond Petaflop/s. Access to the infrastructure is granted on the basis of strict scientific quality through a pan-European peer review system under the guidance of PRACE’s scientific steering committee. Provision of computer time through PRACE started in August 2010 on the supercomputer JUGENE at Research Centre Jülich, Europe’s first Petaflop-system. In this talk I will, in my function as the coordinator of the PRACE development project, describe the design, structure, implementation, purpose and further development of PRACE and I am going to present selected leading edge simulations, as carried out on the 294912-core Petaflop-system at the Jülich Supercomputing Centre.