Reliable and Energy Efficient Streaming Multiprocessor Systems by Anup Kumar Das Akash Kumar Bharadwaj Veeravalli & Francky Catthoor
Author:Anup Kumar Das, Akash Kumar, Bharadwaj Veeravalli & Francky Catthoor
Language: eng
Format: epub
Publisher: Springer International Publishing, Cham
Thus far, the validation of the proposed temperature model is presented. In the next few subsections, results to validate the proposed approach are presented.
4.6.5 MTTF Improvement Considering Task Remapping
Modern multiprocessor systems support remapping of tasks (actors in the SDFG terminology) from faults cores to other working cores. MTTF for these systems need to be computed by considering task remapping, as opposed to the naive way of considering the time to first failure. To highlight the MTTF differences in the two computation techniques, experiments are conducted on a multiprocessor system with six cores and a set of six real-life applications. This is shown in Fig. 4.11a. There are two bars for every application. The left bar is for the MTTF computed considering the first failure and the right one for MTTF computed considering task remapping. As seen from the figure, the two MTTF values are similar for FFT and MP3 decoder applications. For the four other applications, the two MTTF values differ. On average for all applications considered, the MTTF improvement is 15%. To give more insight on the reason for such low MTTF difference for applications such as FFT, as opposed to say, JPEG decoder, Fig. 4.11b plots the mean and the standard deviation of the aging of the different cores for the six applications. The standard deviation of the aging values is a measure of how much the aging of individual cores differ from the mean. A low standard deviation indicates a balanced wear-out of the cores. On the other hand, a high standard deviation indicates some cores age faster than others. The standard deviation is normalized with respect to the mean value of the aging.
Fig. 4.11MTTF difference considering time to first failure and task remapping. (a) MTTF underestimation. (b) Mean and standard deviation of aging
Download
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.
The Mikado Method by Ola Ellnestam Daniel Brolund(26280)
Hello! Python by Anthony Briggs(25206)
Secrets of the JavaScript Ninja by John Resig Bear Bibeault(24435)
Kotlin in Action by Dmitry Jemerov(23526)
The Well-Grounded Java Developer by Benjamin J. Evans Martijn Verburg(22869)
Dependency Injection in .NET by Mark Seemann(22658)
OCA Java SE 8 Programmer I Certification Guide by Mala Gupta(21420)
Algorithms of the Intelligent Web by Haralambos Marmanis;Dmitry Babenko(20260)
Grails in Action by Glen Smith Peter Ledbrook(19332)
Adobe Camera Raw For Digital Photographers Only by Rob Sheppard(17047)
Sass and Compass in Action by Wynn Netherland Nathan Weizenbaum Chris Eppstein Brandon Mathis(16358)
Secrets of the JavaScript Ninja by John Resig & Bear Bibeault(14071)
Test-Driven iOS Development with Swift 4 by Dominik Hauser(12246)
Jquery UI in Action : Master the concepts Of Jquery UI: A Step By Step Approach by ANMOL GOYAL(11520)
A Developer's Guide to Building Resilient Cloud Applications with Azure by Hamida Rebai Trabelsi(10637)
Hit Refresh by Satya Nadella(9212)
The Kubernetes Operator Framework Book by Michael Dame(8574)
Exploring Deepfakes by Bryan Lyon and Matt Tora(8424)
Robo-Advisor with Python by Aki Ranin(8366)