Off-campus UMass Amherst users: To download campus access theses, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this thesis through interlibrary loan.

Theses that have an embargo placed on them will not be available to anyone until the embargo expires.

Document Type

Open Access

Degree Program

Electrical & Computer Engineering

Degree Type

Master of Science in Electrical and Computer Engineering (M.S.E.C.E.)

Year Degree Awarded

2010

Month Degree Awarded

February

Keywords

recovery, AVF, DVFS, monitor network, reliability stabilization

Abstract

For future multicores, a dedicated interconnect subsystem for on-chip monitors was found to be highly beneficial in terms of scalability, performance and area. In this thesis, such a monitor network (MNoC) is used for multicores to support selective error identification and recovery and maintain target chip reliability in the context of dynamic voltage and frequency scaling (DVFS). A selective shared memory multiprocessor recovery is performed using MNoC in which, when an error is detected, only the group of processors sharing an application with the affected processors are recovered. Although the use of DVFS in contemporary multicores provides significant protection from unpredictable thermal events, a potential side effect can be an increased processor exposure to soft errors. To address this issue, a flexible fault prevention and recovery mechanism has been developed to selectively enable a small amount of per-core dual modular redundancy (DMR) in response to increased vulnerability, as measured by the processor architectural vulnerability factor (AVF). Our new algorithm for DMR deployment aims to provide a stable effective soft error rate (SER) by using DMR in response to DVFS caused by thermal events. The algorithm is implemented in real-time on the multicore using MNoC and controller which evaluates thermal information and multicore performance statistics in addition to error information. DVFS experiments with a multicore simulator using standard benchmarks show an average 6% improvement in overall power consumption and a stable SER by using selective DMR versus continuous DMR deployment.

First Advisor

Russell G Tessier

Share

COinS