Scientific computing in the Americas: the challenge of massive parallelism; Valparaiso, Chile; January 3-14, 2011

点击翻译按钮获取中文摘要

Scientific computing as a means of discovery has the promise of solving some of the most enduring challenges in science. Understanding global climate, unraveling the mysteries of turbulence, predicting natural disasters such as earthquakes and tsunamis, untangling the complexities of life by simulating the proteins that make its building- blocks, all these, and more, are being tackled by scientists today using high-performance computing (HPC). And today, more than ever, an opportunity presents itself for exciting discoveries through HPC, thanks to the advent of new computer architectures that bring huge increases in performance, at lower power consumption and much reduced prices. The new HPC hardware based on graphics processors (GPUs) is a technology that can level the playing field for scientists in many parts of the world to participate in leading-edge computational research. This Pan-American Advanced Studies Institutes (PASI) award, jointly supported by the NSF and the Department of Energy (DOE), will take place January 3-14, 2011 at the Universidad Tecnica Federico Santa María (UTFSM) in Valparaíso, Chile. Organized by Dr. Lorena Barba of Boston University, an interdisciplinary group of collaborators and participants will explore the feasibility of adapting and applying GPU technologies to complex, outstanding problems in the physical sciences. Participants will include a complementary mix of geoscientists, computer scientists, mathematicians, and physicists. Lectures, tutorials and hands-on activities will be focused on modern hardware computing, fundamental algorithms for massively parallel computational applications, and recent developments in open source libraries and programming. PASI participants will include approximately 30 U.S. and Latin American students supported by this award and a similar number of local participants supported by Chilean agencies. An ancillary mentoring session for postdoctoral scholars will provide advice on succeeding in modern academic environments. This PASI aims to initiate a pole of development for high-performance computing using modern hardware, and to stimulate international collaboration and future initiatives leading to student exchanges, joint projects and new funding opportunities for the participants. Graduate students and postdoctoral researchers will have a unique learning experience; they will be trained in the use of the most important software tools for HPC, including those specific to GPUs, and the syllabus will put emphasis on the open-source model for science. The chosen applications, e.g., simulation of tsunamis, will attract the attention of the general public, offering opportunities for dissemination in the press and other media. Extensive dissemination of the learning materials online will ensure impact beyond the duration of the institute itself.