A transparent non-invasive file data model for algorithmic skeletons

Abstract

A file data model for algorithmic skeletons is proposed, focusing on transparency and efficiency. Algorithmic skeletons correspond to a high-level programming model that takes advantage of nestable programming patterns to hide the complexity of parallel/distributed applications.Transparency is achieved using a workspace factory abstraction and the proxy pattern to intercept calls on File type objects. Thus allowing programmers to continue using their accustomed programming libraries, without having the burden of explicitly introducing non-functional code to deal with the distribution aspects of their data.A hybrid file fetching strategy is proposed (instead of lazy or eager), that takes advantage of annotated functions and pipelined multithreaded interpreters to transfer files in-advance or on-demand. Experimentally, using a BLAST skeleton application, it is shown that the hybrid strategy provides a good tradeoff between bandwidth usage and CPU idle time.

PList-based Divide and Conquer Parallel Programming

“…Many of these parallel frameworks are implemented in C++, but Java is becoming popular in the High Performance Computing area [11]. Examples of parallel programming frameworks that implement algorithmic skeletons are: Lithium [12], Calcium [13] and Skandium [14].…”

“…• n is a power of two -P owerList definition, • n is a prime number -simple sequential list definition, • n is a product of different factors -P List definition. The function root : N → C applied to n returns the principal nth order unity root: (13) Case n = 2 k A formula that computes the polynomial value in w j (w j = (root.n) j ) is:…”

“…8. Repeated Rectangle Formula: execution time for data size 2 13 and various parallel recursion depths -Bounded Parallel execution vs. Unbounded Parallel execution repeated rectangle formula (eq. 8-9).…”

A survey of algorithmic skeleton frameworks: high‐level structured parallel programming enablers

“…Second, a type system for nestable skeletons implemented using Java Generics 60. Third, a transparent algorithmic skeleton file access model for data‐intensive applications 74.…”

A Skandium Based Parallelization of DNSSEC

“…In this paper we show how a Java [27] based algorithmic skeleton library, Skandium [29], can be used to parallelize DNSSEC's digital signing of domain zones. Skandium is a complete reimplementation of Calcium [10,11,9] (part of the ProActive Middleware [8]), but focused on parallel multi-core computing. This paper is organized as follows.…”