Architectural support for synchronization-free deterministic parallel programming

Segulja, Cedomir; Abdelrahman, Tarek S.

doi:10.1109/hpca.2012.6169038

Cited by 9 publications

(1 citation statement)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Task superscalar [12] uses memory renaming to eliminate output dependencies in task-based execution. ROKO [29] maintains serial semantics using read and write timestamps for every shared address, but only a single version of the data is kept. While simplifying versioning implementation, this does not eliminate output and anti-dependencies, and requires that read operations update a versioning data structure.…”

Section: Limitations Of Existing Memory Interfacesmentioning

confidence: 99%

O-structures

Gilad

Mackay

Oskin

et al. 2014

Proceedings of the Workshop on Memory Systems Performance and Correctness

View full text Add to dashboard Cite

This paper introduces O-structures, a novel architectural memory element that can be used to facilitate parallelism in task-based execution models. Much like register renaming, each write to an O-structure creates a new version of program memory at that location. These versions can be accessed concurrently and out of program order. O-structures provide a set of semantics that match the needs of task-based execution models, specifically allowing tasks to synchronize on specific versions of memory as well as coordinate access when the necessary version is not known at compile time.In this work, we describe O-structures and provide their complete semantics. We also discuss how a task-based execution of basic data structure manipulations on common data structures (arrays, lists, trees, etc) operate. Results are presented that measure the exposed memory-level parallelism (MLP) in these operations. We find that for previously difficult to parallelize data-structures, such as linked lists, binary trees and sparse-matrix codes we see significant memory level parallelism (50-100 operations per cycle) when using O-structures.

show abstract

Section: Limitations Of Existing Memory Interfacesmentioning

confidence: 99%