Data-Intensive Distributed Systems LaboratoryCFP (TXT, PDF) | News | Topics | Dates | Submission | Organization | Program | Sponsors
7th Workshop on Many-Task Computing on Clouds, Grids, and Supercomputers (MTAGS) 2014
Co-located with Supercomputing/SC 2014In cooperation with ACM SIGHPC
New Orleans, Louisiana -- November 16th, 2014
Keynote
- Fellow at University of Chicago
- Software Architect at Argonne National Laboratory
Dataflow approaches to many-task computing: past, present and future
Abstract: Dataflow-driven programming and execution models can trace their origins to advanced LISP work in the 1970's and some efforts that were even earlier. Since then, dataflow has been frequently revisited - in large part because of the belief that it holds the key to transparent, easier parallelism. We trace currently active work on this problem back to its roots, assess the current state-of-the-art, and suggest areas for future research in solving problems in many-task applications at scales ranging from the personal to the extreme.
Michael
Wilde is a software architect in the Mathematics and Computer
Science Division, Argonne National Laboratory, and a Senior Fellow of
the University of Chicago/Argonne National Laboratory Computation
Institute. His research focus is the application of parallel scripting
to enhance scientific productivity by making parallel and distributed
computing systems easier to use. He also conducts research into data
provenance to record and query the history and metadata of scientific
computations and datasets. His work centers on development and
application of the Swift parallel scripting language,
http://swift-lang.org.