DocumentCode
2253742
Title
Random pattern testing for MCM interconnect using computer generated hologram technology
Author
Chang, Chi-yuan
Author_Institution
Dept. of Comput. Sci. & Inf. Eng., Jinwen Univ. of Sci. Technol., Taipei, Taiwan
Volume
5
fYear
2010
fDate
11-14 July 2010
Firstpage
2653
Lastpage
2657
Abstract
Future computers will need to incorporate the parallelism of optical interconnections in order to achieve projected performance within reasonable size, power, and speed. Optical interconnection provides more advantage than traditional electrical interconnection in large system. In this paper we will propose a Markov model to determine the number of random patterns required for testing of optical interconnect. We describe the random vector to test interconnect and show that a little of random vectors can give close to 100% detection quality. The faults under consideration include stuck-at faults and bridging (short) faults. The simulation results suggest that if the length of boundary scan is large, the traditional method on minimization of test vectors may not lead to a better result. In stead, one should concentrate on choosing the random vector.
Keywords
computer-generated holography; optical interconnections; MCM interconnect; Markov model; boundary scan; bridging faults; computer generated hologram; optical interconnections; random pattern testing; random vector; stuck-at faults; Computers; Driver circuits; Integrated circuit interconnections; Machine learning; Markov processes; Optical interconnections; Testing; Interconnects; Markov model; Random Testing;
fLanguage
English
Publisher
ieee
Conference_Titel
Machine Learning and Cybernetics (ICMLC), 2010 International Conference on
Conference_Location
Qingdao
Print_ISBN
978-1-4244-6526-2
Type
conf
DOI
10.1109/ICMLC.2010.5580914
Filename
5580914
Link To Document