Novel runtime systems support for adaptive compositional modeling in PSEs

Abstract

Grid infrastructures and computing environments have progressed significantly in the past few years. The vision of truly seamless Grid usage relies on runtime systems support that is cognizant of the operational issues underlying grid computations and, at the same time, is flexible enough to accommodate diverse application scenarios. This paper addresses the twin aspects of Grid infrastructure and application support through a novel combination of two computational technologies -Weaves, a source-language independent parallel runtime compositional framework that operates through reverse-analysis of compiled object files, and runtime recommender systems that aid in dynamic knowledge-based application composition. Domain-specific adaptivity is exploited through a novel compositional system that supports runtime recommendation of code modules and a sophisticated checkpointing and runtime migration solution that can be transparently deployed over Grid infrastructures. A core set of "adaptivity schemas" are provided as templates for adaptive composition of large-scale scientific computations. Implementation issues, motivating application contexts, and preliminary results are described.

The Adaptive Code Kitchen: Flexible Tools for Dynamic Application Composition

“…These applications include the Sweep3D benchmark for discrete-ordinates neutron transport, collaborating ELLPACK partial differential equation solvers, and checkpoint and run-time migration of parallel grid applications. For more details refer to [22,34].…”

“…Supported in part by NSF CAREER grant EIA-9984317, this idea has been recently extended in many important ways -mining recommendation spaces with continuous-valued attributes, a recommendation portal with database and experiment management support for performance data analysis, and automatic mining of recommendation spaces, providing support for algorithm selection at runtime, knowledge-based compositional modeling, and harnessing domain knowledge of physical properties underlying problems. See [34] for more details.…”

“…A schema thus suggests a high-level mode of problem solving [3], not unlike parameter sweeps [10] and algorithmic bombardment [6], and is pertinent in specific application contexts. A selection of adaptivity schemas has been presented elsewhere [34] and hence we do not elaborate on these further in this paper.…”

“…In dynamic computations (moving grids or changing physics) the characteristics of the linear system may also change with time. We plan to materialize the adaptive schema skeletons from [34] to address all the above opportunities and use the realized compositions to simulate leading edge film cooling flows for gas turbine blades.…”

“…We envision adaptive composition [34] as supporting componentization over native object codes (language neutrality), providing arbitrary code expansion, contraction, and substitution operations (flexible adaptation), and incorporating runtime recommender systems for algorithm selection (model-based control). Existing solutions typically emphasize one or at most two of these aspects and do not provide an integrated environment to explore adaptivity scenarios over arbitrary scientific codes.…”