DocumentCode
2644455
Title
An Enhanced Strategy for Functional Stress Pattern Generation for System-on-Chip Reliability Characterization
Author
De Carvalho, M. ; Bernardi, P. ; Sanchez, E. ; Reorda, M. Sonza
Author_Institution
Dipt. di Autom. e Inf., Politec. di Torino, Torino, Italy
fYear
2010
fDate
13-15 Dec. 2010
Firstpage
29
Lastpage
34
Abstract
Reliability characterization is the industrial process intended to measure the useful life period and failure rate of a component population by exploiting stress mechanisms. The paper describes a methodology for the automatic generation of stress programs to be used during the reliability characterization process of Systems-on-Chip (SoC). The proposed methodology is composed of a two-phase strategy, first an evolutionary algorithm (EA) works on the SoC´s description at Register-Transfer-Level (RTL) by evaluating high-level metrics to quickly progress to a sufficient stress quality level, then evolution is continued on the gate-level description towards a better stress quality. The proposed methodology was experimented on a SoC manufactured in a 90nm technology including an 8051 processor. The proposed strategy reduces significantly the generation times and quickly improves the stress quality values with respect to the previous methodology.
Keywords
evolutionary computation; integrated circuit reliability; system-on-chip; 8051 processor; SoC; SoC description; evolutionary algorithm; functional stress pattern generation; gate-level description; high-level metrics evaluation; industrial process; register-transfer-level; reliability characterization process; size 90 nm; stress quality level; systems-on-chip; two-phase strategy; Bismuth; Logic gates; Measurement; Reliability; Stress; Switches; System-on-a-chip; Functional Stress Pattern Generation; SoC Reliability Characterization; Stress;
fLanguage
English
Publisher
ieee
Conference_Titel
Microprocessor Test and Verification (MTV), 2010 11th International Workshop on
Conference_Location
Austin, TX
ISSN
1550-409
Print_ISBN
978-1-61284-287-5
Type
conf
DOI
10.1109/MTV.2010.14
Filename
5976240
Link To Document