The Charm++ Workshop is aimed both at taking stock of the collaborative interdisciplinary research projects in which the Parallel Programming Laboratory (the developers of Charm++ and AMPI) participates, and exploring on-going and developing research areas which employ the primary tools provided by PPL.

Our group's goal is to develop technology that improves performance of parallel applications while also improving programmer productivity. We aim to reach a point where, with our freely distributed software base, complex irregular and dynamic applications can (a) be developed quickly and (b) perform scalably on machines with thousands of processors.

In addition to the computer scientists at PPL, the collaborators represented at this workshop include those from University of Illinois at Urbana Champaign (CS, ME, TAM, MS, Physics, Beckman), University of Washington Seattle, Princeton University, NYU, IBM and Pittsburgh Supercomputing Center (PSC).

• 

  Panel Discussion: Exascale by 2018. Really?

  Wednesday, April 28th: 4:30 pm - 6:00 pm (Tentative)

  An open panel will discuss the feasibility of delivering exascale performance by 2018. Can it be done? If not by 2018 then when? Panelists will take opposing viewpoints and reveal what challenges the supercomputing community still faces before the launch of the first exascale machine.
• 

  Blue Waters Facility Tour

  Friday, April 30th: 12:00 pm - 1:00 pm (Tentative)

  Blue Waters is expected to be the fastest supercomputer for open scientific research when it launches in 2011. Blue Waters will be operated on the University of Illinois-Urbana Champaign campus and workshop participants will have the chance to tour the facilities built for the massive supercomputer. More information about the Blue Waters project can be found on the NCSA website.
• 

  Heterogeneous Computing with Charm++

  Accelerators such as Graphical Processing Units (GPUs) and specialized cores, such as the Synergistic Processing Elements (SPEs) on the Cell processor, are being used with greater frequency in the realm of parallel computing to speedup computationally heavy portions of code. These systems are comprised of multiple types of processing elements, each with unique characteristics, strengths, weaknesses, and programming paradigms. Developing applications can be challenging since many architectural details must be taken into account.
  
  

  Related talks:

  • Charm++ on Heterogeneous Systems
  • Studying the Scalability of N-body Simulations on GPU Clusters
• 

  NAMD: Biomolecular Simulation

  NAMD is a portable parallel application for biomolecular simulations. NAMD pioneered the use of hybrid spatial and force decomposition, a technique used now by most scalable programs for biomolecular simulations, including Blue Matter and Desmond developed by IBM and D. E. Shaw respectively. NAMD is developed using Charm++ and benefits from its adaptive communication-computation overlap and dynamic load balancing.
  
  

  Related talks:

  • Preparation for Blue Waters
  • Charm++ Hits and Misses
  • Hierarchical Load Balancers