Skip to main content

Efficient Data Supply for Parallel Heterogeneous Architectures

Author(s): Ham, Tae J; Aragón, Juan L; Martonosi, Margaret

To refer to this page use:
Full metadata record
DC FieldValueLanguage
dc.contributor.authorHam, Tae J-
dc.contributor.authorAragón, Juan L-
dc.contributor.authorMartonosi, Margaret-
dc.identifier.citationHam, Tae Jun, Juan L. Aragón, and Margaret Martonosi. "Efficient Data Supply for Parallel Heterogeneous Architectures." ACM Transactions on Architecture and Code Optimization (TACO) 16, no. 2 (2019): 9:1-9:23. doi:10.1145/3310332en_US
dc.description.abstractDecoupling techniques have been proposed to reduce the amount of memory latency exposed to high-performance accelerators as they fetch data. Although decoupled access-execute (DAE) and more recent decoupled data supply approaches offer promising single-threaded performance improvements, little work has considered how to extend them into parallel scenarios. This article explores the opportunities and challenges of designing parallel, high-performance, resource-efficient decoupled data supply systems. We propose Mercury, a parallel decoupled data supply system that utilizes thread-level parallelism for high-throughput data supply with good portability attributes. Additionally, we introduce some microarchitectural improvements for data supply units to efficiently handle long-latency indirect loads.en_US
dc.format.extent9:1 - 9:23en_US
dc.relation.ispartofACM Transactions on Architecture and Code Optimizationen_US
dc.rightsFinal published version. This is an open access article.en_US
dc.titleEfficient Data Supply for Parallel Heterogeneous Architecturesen_US
dc.typeJournal Articleen_US

Files in This Item:
File Description SizeFormat 
EfficientDataSupplyParallelArchitectures.pdf1.79 MBAdobe PDFView/Download

Items in OAR@Princeton are protected by copyright, with all rights reserved, unless otherwise indicated.