Reduction Of Architecture Vulnerability Factor Using Modified Razor Flipflops

Simple item page
dc.contributor.advisorFranklin, Manojen_US
dc.contributor.authorSeshadri, Kiran Kalkunteen_US
dc.contributor.departmentElectrical Engineeringen_US
dc.contributor.publisherDigital Repository at the University of Marylanden_US
dc.contributor.publisherUniversity of Maryland (College Park, Md.)en_US
dc.date.accessioned2008-04-22T16:09:43Z
dc.date.available2008-04-22T16:09:43Z
dc.date.issued2007-12-17en_US
dc.description.abstractResearch has shown that microprocessors and structures of the microprocessors are vulnerable to alpha Single Event Upsets that affect program correctness and reliability. In this thesis, we have explored the use of Modified Razor flip-flops in the microprocessor to increase the overall reliability of the microprocessor. We have adopted Architecturally Correct Execution (ACE) time based techniques to measure the Architecture Vulnerability Factor (AVF) of high performance microprocessors and their internal structures using the SPEC 2000 integer benchmarks. We have computed the reduction in AVF with the introduction of Modified Razor flip-flops for various combinations of bit-fields that have high vulnerability. However, introduction of Modified Razor flip-flops results in higher area requirement on the die and higher power consumption. We have identified the most cost-effective solution by identifying the fields of these microarchitectural structures - where Modified Razor flip-flops are introduced - that result in the highest percentage decrease in AVF per unit area-power product.en_US
dc.format.extent3530191 bytes
dc.format.mimetypeapplication/pdf
dc.identifier.urihttp://hdl.handle.net/1903/7829
dc.language.isoen_US
dc.subject.pqcontrolledEngineering, Electronics and Electricalen_US
dc.subject.pqcontrolledEngineering, Electronics and Electricalen_US
dc.titleReduction Of Architecture Vulnerability Factor Using Modified Razor Flipflopsen_US
dc.typeThesisen_US

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
umi-umd-5115.pdf
Size:
3.37 MB
Format:
Adobe Portable Document Format
Download

Collections

UMD Theses and Dissertations
Electrical & Computer Engineering Theses and Dissertations