"GPUs have evolved to the point where many real world applications are easily implemented on them and run significantly faster than on multi-core systems. Future computing architectures will be hybrid systems with parallel-core GPUs working in tandem with multi-core CPUs." -- Prof. Jack Dongarra, Director, Innovative Computing Laboratory, University of Tennessee.
The Application Acceleration Center of Excellence (AACE) was established by the Joint Institute for Computational Sciences (JICS) in 2011 in partnership with NICS, Cray, AMD, and other institutions including the University of Tennessee, Virginia Polytechnic Institute and State University, Georgia Institute of Technology, Florida State University, University of California Santa Barbara, Indiana University, and Jackson State University. Its twofold mission is to optimize applications for current and future accelerator-based compute systems and to develop expertise in the expression and exploitation of fine-grain and medium-grain parallelism crucial for exascale computing, making AACE an essential element of a sustainable software infrastructure for simulation in science and engineering.
The center’s objectives are:
- Accelerate NSF projects toward exascale computing by exploiting state-of-the-art heterogeneous architectures
- Engage the broader community in the development of new algorithms and science codes optimized for accelerator-based architectures
- Disseminate fundamental knowledge and principles related to the effective and efficient use of accelerators through training courses, educational materials, and research publications
- Host visiting students and faculty at NICS to facilitate effective exchange of expertise and cross-disciplinary collaboration
Related projects:
- The Keeneland Project is a five-year, $12 million Track 2D grant awarded by the NSF for the deployment of an experimental high performance system. The Georgia Institute of Technology (Georgia Tech) and its partners, the University of Tennessee at Knoxville and the Oak Ridge National Laboratory will initially acquire and deploy a small, experimental, high-performance computing system consisting of an HP system with NVIDIA Tesla accelerators attached. The project team will use this initial system to develop scientific libraries and programming tools to facilitate the development of science and engineering research applications, while also providing consulting support to researchers who wish to develop applications for the system. In 2012, the project will upgrade the heterogeneous system to a larger and more powerful system based on a next-generation platform and NVIDIA accelerators. It is anticipated that the final system will have a peak performance of roughly 2 petaflops. The project will then operate the upgraded system as a TeraGrid resource for two years.
- The Center for Remote Data Analysis and Visualization (RDAV) operated by the University of Tennessee is sponsored by the National Science Foundation as part of the TeraGrid XD (eXtreme Digital) project. Its purpose is to aid in the significant challenge of transforming ever-increasing amounts of digital data into knowledge and insight by providing scientists with well-engineered and well-supported remote visualization, analysis, and scientific workflow technologies.
