Slides available here!

Speaker:

Prof. Norbert Wehn, University of Kaiserslautern, Germany

Title:

An Embedded Computing Architecture for Finding Similarities in Large Networks

Abstract:

A very important task in many big data applications like recommendation systems, customer behavior modelling or incident detection systems is the extraction of network motifs in very large networks. In this talk, we will present an embedded 28nm ASIC architecture for computing similarities in large networks based on the co-occurrence network motif. We use a Netflix dataset consisting of 100.480.507 user ratings for 17.700 movies to highlight the performance of our new embedded architecture. Our new architecture outperforms a 10 node Intel cluster in speed, each node consisting of 2 Intel Xeon X5680, and by 44x with respect to memory requirements and about 200x faster with an equivalent power consumption.

Bio:

Norbert holds the chair for Microelectronic System Design in the department of Electrical Engineering and Information Technology at the University of Kaiserslautern. He has more than 250 publications in various fields of microelectronic system design and holds several patents. Two start-ups spinout of his research group. He is Vice-President of the University Kaiserslautern, associate editor of various journals and member of several scientific advisory boards. In 2003 he served as program chair for DATE 2003 and as general chair for DATE 2005 respectively. In 2014 he was general Co-Chair of FPL 2015. His special research interests are VLSI-architectures for mobile communication, forward error correction techniques, low-power techniques, advanced SoC architectures, 3D integration, reliability issues in SoC and hardware accelerators for financial mathematics and big data applications.



* If you wish to modify any information or update your photo, please contact the web chairmpsoc2014@imag.fr