Show simple item record

dc.contributor.advisorHill, Richarden
dc.contributor.advisorBagdasar, Ovidiuen
dc.contributor.advisorJones, Cliftonen
dc.contributor.authorKainth, Haresh S.
dc.date.accessioned2014-02-25T17:04:42Zen
dc.date.available2014-02-25T17:04:42Zen
dc.date.issued2014-01-29en
dc.identifier.urihttp://hdl.handle.net/10545/313343en
dc.description.abstractMulticore processors often increase the performance of applications. However, with their deeper pipelining, they have proven increasingly difficult to improve. In an attempt to deliver enhanced performance at lower power requirements, semiconductor microprocessor manufacturers have progressively utilised chip-multicore processors. Existing research has utilised a very common technique known as thread-level speculation. This technique attempts to compute results before the actual result is known. However, thread-level speculation impacts operation latency, circuit timing, confounds data cache behaviour and code generation in the compiler. We describe an software framework codenamed Lyuba that handles low-level data hazards and automatically recovers the application from data hazards without programmer and speculation intervention for an asymmetric chip-multicore processor. The problem of determining correct execution of multiple threads when data hazards occur on conventional symmetrical chip-multicore processors is a significant and on-going challenge. However, there has been very little focus on the use of asymmetrical (heterogeneous) processors with applications that have complex data dependencies. The purpose of this thesis is to: (i) define the development of a software framework for an asymmetric (heterogeneous) chip-multicore processor; (ii) present an optimal software control of hardware for distributed processing and recovery from violations;(iii) provides performance results of five applications using three datasets. Applications with a small dataset showed an improvement of 17% and a larger dataset showed an improvement of 16% giving overall 11% improvement in performance.
dc.language.isoenen
dc.publisherUniversity of Derbyen
dc.subjectThread-level speculationen
dc.subjectTLSen
dc.subjectMulticoreen
dc.subjectIBM cellen
dc.subjectProgrammingen
dc.subjectData hazardsen
dc.subjectAsymmetric architectureen
dc.subjectHeterogenousen
dc.titleA data dependency recovery system for a heterogeneous multicore processoren
dc.typeThesis or dissertationen
dc.type.qualificationnamePhDen
dc.type.qualificationlevelDoctoralen
refterms.dateFOA2019-02-28T13:24:13Z
html.description.abstractMulticore processors often increase the performance of applications. However, with their deeper pipelining, they have proven increasingly difficult to improve. In an attempt to deliver enhanced performance at lower power requirements, semiconductor microprocessor manufacturers have progressively utilised chip-multicore processors. Existing research has utilised a very common technique known as thread-level speculation. This technique attempts to compute results before the actual result is known. However, thread-level speculation impacts operation latency, circuit timing, confounds data cache behaviour and code generation in the compiler. We describe an software framework codenamed Lyuba that handles low-level data hazards and automatically recovers the application from data hazards without programmer and speculation intervention for an asymmetric chip-multicore processor. The problem of determining correct execution of multiple threads when data hazards occur on conventional symmetrical chip-multicore processors is a significant and on-going challenge. However, there has been very little focus on the use of asymmetrical (heterogeneous) processors with applications that have complex data dependencies. The purpose of this thesis is to: (i) define the development of a software framework for an asymmetric (heterogeneous) chip-multicore processor; (ii) present an optimal software control of hardware for distributed processing and recovery from violations;(iii) provides performance results of five applications using three datasets. Applications with a small dataset showed an improvement of 17% and a larger dataset showed an improvement of 16% giving overall 11% improvement in performance.


Files in this item

Thumbnail
Name:
A Data Dependency Recovery System ...
Size:
3.296Mb
Format:
PDF
Description:
Converted to Standard PDF format

This item appears in the following Collection(s)

Show simple item record