Summer School 2010 Workshops

1. Scaling to Petascale
   July 6–9, 2010 (3–4 days)
   Prerequisite: MPI
   
• Opening Keynote: High-performance Computing at the Petascale
• Hybrid Programming Models
• MPI + OpenMP
• Partitioned Global Address Space (PGAS) Languages
  
• Unified Parallel C
• Co-Array Fortran
• Libraries (e.g., PETSc, SPRNG, pnetCDF)
  
• Load Balancing
  
• Parallel I/O
• Visualization Tools (e.g., VisIt)
  
• Tools:
  
• IDEs (e.g., Eclipse)
• Debugging Tools (e.g., DDT)
• Performance Tools
• Compiler Optimization
• Case Studies:
• Cosmology (http://www.ncsa.illinois.edu/News/Stories/dark_days/)
  
• Hands-on Lab
• Using Eclipse
• OpenMP Coding
  
• Closing Keynote: The Path to Exascale
  
2. Big Data for Science
   July 26–30, 2010 (3–5 days)
   Prerequisite: any high-level programming language such as C, C++, or Java
   
• Opening Keynote: Data-intensive Computing
• Data Movement & Storage
• Relational Databases & Database Frameworks
• Oracle Real Application Clusters
  
• Amazon RDS
  
• SciDB
  
• Keynote: Distributed Data-Parallel Computing
• MapReduce for Big Data
• Hybrid Approaches to Big Data (e.g., HadoopDB, Sector/Sphere)
• Performance Considerations
  
• Visualization of Large Data Sets
  
• Case Studies:
• Astronomy
• Bioinformatics
• Earth Science
  
• Hands-on Lab
3. Many-Core Processors
   August 2–6, 2010 (3–5 days)
   Prerequisite: C or C++
   
• Opening Keynote: Heterogeneous Computing for Performance
  
• Intro to CUDA
• CUDA Threads & Memory
• CUDA Tools & Libraries
• Intro to OpenCL
• OpenCL Platform Model
• OpenCL Execution Model
• OpenCL Memory Model
• Hybrid Programming Models
• MPI + CUDA
• MPI + OpenCL
  
• Performance Considerations
• Case Studies:
• Medical Imaging (Utah)
  
• Gene Sequencing (Maryland)
• Molecular Dynamics (http://www.ncsa.illinois.edu/News/Stories/get_game/)
• Computational Neuroscience
  
• Hands-on Lab