State access patterns in stream parallel computations. (November 2018)

• Record Type: Journal Article
• Authors: Danelutto, Marco; Kilpatrick, Peter; Mencagli, Gabriele; Torquati, Massimo
• Other Names: Balaji Pavan guest-editor.; Leung Kai-Cheung guest-editor.; Huang Zhiyi guest-editor.; García-Blas Javier guest-editor.; Brown Christopher guest-editor.
• Is Part Of: International journal of high performance computing applications. Volume 32:Number 6(2018)
• Page Start: 807
• Page End: 818
• Publication Date: 2018-11
• Subjects: Structured parallel programming -- parallel design patterns -- stateful computations; High performance computing -- Periodicals; Supercomputers -- Periodicals; 004.1105
• DOI: 10.1177/1094342017694134
• Languages: English
• ISSNs: 1094-3420

Abstract

We introduce a set of state access patterns suitable for managing accesses to states in parallel computations operating on streams. The state access patterns are useful for modelling typical stream parallel applications. We present a classification of the patterns according to the extent and way in which the state can be structured and accessed. We define precisely the state access patterns and discuss possible implementation schemas, performances and possibilities to manage adaptivity (parallelism degree) in the patterns. We present experimental results relative to implementations built on top of the structured parallel programming framework FastFlow that demonstrate the feasibility and efficiency of the proposed access patterns.