APGAS09
IBM Research Homepage 
 Research Home  >>  Workshop on Asynchrony in the PGAS Programming Model


23rd International Conference on Supercomputing


Call for Papers


Paper submission for APGAS09


PGAS Languages

PGAS@IBM


HPC Challenge Awards website




Workshop on Asynchrony in the PGAS Programming Model
June 8, 2009
Collocated with ICS 2009

The new golden era of concurrency offers a diversity of concurrent architectures to the application programmer -- clusters of symmetric multiprocessors, heterogeneous accelerators (such as the Cell, GPGPUs), large core-count integrated machines (such as the Blue Gene), and multiple levels of parallelism (multithreaded multicores, such the newer Power architectures, or distributed multiprocessors like the Cray MTA).

The central programming challenge for these new architectures is to develop a suitable and robust programming model. The Partitioned Global Address Space (PGAS) model has attracted considerable attention as such a programming model, particularly in the High Performance Computing space. However, widespread acceptance and extension to newer paradigms (e.g. to heteregeneous accelerators and irregular application domains) has been hampered by a focus on homogoneous execution contexts and the Single Program Multiple Data (SPMD) threading model.

Several research groups have recently started to investigate asynchronous execution in the PGAS model. Asynchronous execution is the foundation for active message programming, and fine-grained concurrency (e.g. Cilk-style fork-join, OpenMP style teams). It is also the basis for data transfer between non-coherent memories (DMA), remote atomic operations, and in general for techniques for overlapping computation and communication. Finally, asynchronous execution offers a simple model for spawning massive multi-threaded kernel computations on accelerators such as GPUs.

Authors are invited to submit a 10 page paper for consideration for the workshop. We invite papers on all aspects of the PGAS programming model, with particular emphasis on handling asynchronous (multi-threaded) execution (topics enumerated below).

Workshop program

Important dates

Submission deadline Notification of Acceptance Final copy due
Extended to April 27, 2009 May 4, 2009 May 25, 2009

Early conference registration ends on May 10, 2009.

Paper submission here. Final papers will be hosted on this website.

Organizing Committee
George Almasi IBM Research
Calin Cascaval IBM Research
Vijay Saraswat IBM Research

Program Committee
George Almasi IBM Research
Guy Blelloch CMU
Calin Cascaval IBM Research
Brad Chamberlain Cray Inc
Tarek el-Ghazawi (George Washington U)
Kath Knobe Intel
James Larus Microsoft Research
Ewing "Rusty" Lusk Argonne Nat. Lab
Jan-Willem Maessen Sun Microsystems
Robert Numrich U Minnesota
Martin Rinard MIT
Bratin Saha Intel
Vijay Saraswat IBM Research
Lauren Smith NSA
Marc Snir U Illinois at Urbana-Champaign
Kathy Yelick UC Berkeley

Workshop topics

Programming models
  • Reasoning about algorithms implemented in the asynchronous model
  • Determinacy
  • Concurrency refactoring
Language design
  • SPMD PGAS languages (e.g. UPC, Titanium, Co-Array Fortran)
  • APGAS extensions to OpenMP, OpenCL
  • Asynchronous PGAS languages (e.g. X10, Chapel)
  • Type systems (e.g. place-based type systems), type inference in a PGAS context
  • New language constructs
  • Programmer directed code-generation (e.g. sketching, generative language features)
Compilation issues
  • Code generation
  • Concurrency optimizations: locality, synchronization reduction, communication reduction, computation/communication overlap
  • Compilation for heterogeneous architectures
Runtime
  • Algorithmic scheduling (e.g. work-stealing)
  • Auto-tuning
Debugging
  • Scalability of debugging tools
  • Challenges of asynchronous debugging (debugging without barriers)
Libraries and Interoperability
  • Data translation between SPMD and asynchronous paradigms
  • Composability of libraries
  • Mapping APGAS execution concepts to MPI
  • Non-blocking and one-sided collective operations as a way to bridge APGAS and SPMD concepts
  • Parallel I/O

 Privacy | Legal | Contact | IBM Home | Research Home | Project List | Research Sites | Page Contact <gheorghe@us.ibm.com> $Id: apgas09.html,v 1.13 2009/04/21 14:33:15 almasig Exp $