Toward a common component architecture for high-performance scientific computing

Armstrong, Rob; Geist, Al; Keahey, K.; Kohn, Scott R.; McInnes, Lois Curfman; Parker, Steven G.; Smolinski, Brent A.

doi:10.1109/hpdc.1999.805289

Cited by 260 publications

(219 citation statements)

References 17 publications

Supporting

Mentioning

215

Contrasting

Unclassified

Order By: Relevance

“…about past performance of the system) in order to choose the best library and algorithmic strategy for solving the current problem at hand; • A history database that not only records all the information that the intelligent component creates or acquires, but also all the data (e.g., algorithm, hardware, or performance related) that each interaction with a numerical routine produces; • A system component that provides the interface to the available computational resources (whether on a desktop, in a cluster or on a Grid), combining the decision of the intelligent component with both historical information and its own knowledge of available resources in order to schedule the given problem for execution; • A metadata vocabulary that expresses properties of the user data and of performance profiles, and that will be used to build the performance history database. By considering this as behavioural metadata, we are led to intelligent software components as an extension of the CCA [1,11] framework. The metadata associated with user input makes it possible for the user to pass various degrees of information about the problem to be solved.…”

Section: Components Of a Sans Systemmentioning

confidence: 99%

Self-Adapting Numerical Software and Automatic Tuning of Heuristics

Dongarra¹,

Eijkhout²

2003

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Self-Adapting Numerical Software (SANS) systems aim to bridge the knowledge gap that exists between the expertise of domain scientists, and the know-how that is needed to fulfill efficiently their computational demands. This know-how extends to algorith choice, computational grid utilization, and use of properly optimized kernels. A SANS system is a piece of meta software that mediates between the application program and the computational platform so that application scientists -with disparate levels of knowledge of algorithmic and programmatic complexities of the underlying numerical software -can easily realize numerical solvers and efficiently solve their problem.The main component of a SANS system is an Intelligent Agent that automates method selection based on data, algorithm and system attributes. The IA uses heuristics to make its decisions. In this paper we explain how the heuristics of the IA can be tuned over time by redundant testing and using the nature of many applications. IntroductionIn numerous technologically important areas, such as aerodynamics (vehicle design), electrodynamics (semiconductor device design), magnetohydrodynamics (fusion energy device design), and porous media (petroleum recovery), production runs on expensive, high-end systems last for hours or days, and a major portion of the execution time is usually spent inside of numerical routines, such as for the solution of large-scale nonlinear and linear systems that derive from discretized systems of nonlinear partial differential equations. These numerical parts of the code can contain a large number of tuning parameters, the choice of which greatly influences the efficiency of the total code, or can even make the difference between obtaining a solution and obtaining none.Such numerical concerns, however, are artifactual from the perspective of the scientific and engineering users, who are usually more concerned with modeling and discretization issues. The classic response to numerics was to encode the requisite mathematical, algorithmic and programming expertise into libraries that could be easily reused by a broad spectrum of domain scientists. However, in high-performance computing this solution is no longer sufficient. There is typically more than one algorithm for the stated purpose, and since several levels of algorithms are needed in a large-scale application; the different algorithms can have interlocking parameter settings, and the availability of parallel computing platforms influences algorithmic decisions. Since the difference in performance between an optimal choice of algorithm and hardware, and a less than optimal one, can span orders of magnitude, it is unfortunate that selecting the right solution strategy requires specialized knowledge of both numerical analysis and of computing platform characteristics. Our SANS system aims to assist the application user to navigate this maze of computational possibilities. *

show abstract

Section: Components Of a Sans Systemmentioning

confidence: 99%

Self-Adapting Numerical Software and Automatic Tuning of Heuristics

Dongarra¹,

Eijkhout²

2003

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…Hierarchical composition greatly improves the expressive power of the component model and is inherited by GCM from the Fractal component model [8]. -Structured: In addition to standard intra-component interaction mechanisms (use/provide ports [9]) GCM allows components to interact through collective ports modelling common structured parallel computation communication patterns. These patterns include broadcast, multicast, scatter and gather communications operating on collections of components.…”

Section: The Gcm Frameworkmentioning

confidence: 99%

Integrating Autonomic Grid Components and Process-Driven Business Applications

Weigold

Aldinucci

Danelutto

et al. 2010

Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering

View full text Add to dashboard Cite

Abstract. Today's business applications are increasingly process driven, meaning that the main application logic is executed by a dedicate process engine. In addition, component-oriented software development has been attracting attention for building complex distributed applications. In this paper we present the experiences gained from building a process-driven biometric identification application which makes use of Grid infrastructures via the Grid Component Model (GCM). GCM, besides guaranteeing access to Grid resources, supports autonomic management of notable parallel composite components. This feature is exploited within our biometric identification application to ensure real time identification of fingerprints. Therefore, we briefly introduce the GCM framework and the process engine used, and we describe the implementation of the application using autonomic GCM components. Finally, we summarize the results, experiences, and lessons learned focusing on the integration of autonomic GCM components and the process-driven approach.

show abstract

“…The Common Component Architecture (CCA) project [6] is a major research and development project focused on composition of parallel programs from components. One primary goal of CCA is to enable composition of programs from components written in multiple languages.…”

Section: Related Researchmentioning

confidence: 99%

“…Exceptions include SciRun [23,26] and several implementations of variants of the Common Component Architecture [6] including XCAT [19], Ccaffeine [14] and Babel [7,9]. The execution environments for parallel programs are much more diverse than those for sequential programs.…”

Section: Introductionmentioning

confidence: 99%

Compositional Development of Parallel Programs

Mahmood

Deng

Browne

2004

Languages and Compilers for Parallel Computing

View full text Add to dashboard Cite

Abstract. This paper presents a programming model, a language (P-COM 2 ) and a compiler that composes parallel and distributed programs from independently written components. P-COM 2 is an interface definition language which incorporates information on behaviors and implementations of components to enable qualification of components for effectiveness in specific application instances and execution environments. The programming model targets development of families of related programs. One objective is to be able to compose programs which are near-optimal for given application instances and execution environments. Component-oriented development is motivated for parallel and distributed computations. The programming model is defined and described and illustrated with a simple example. The compilation process is briefly defined and described. Experience with one more complex application, a generalized fast multipole solver is sketched including performance data, some of which was surprising.

show abstract

Toward a common component architecture for high-performance scientific computing

Cited by 260 publications

References 17 publications

Self-Adapting Numerical Software and Automatic Tuning of Heuristics

Self-Adapting Numerical Software and Automatic Tuning of Heuristics

Integrating Autonomic Grid Components and Process-Driven Business Applications

Compositional Development of Parallel Programs

Contact Info

Product

Resources

About