ABSTRACT
Surprisingly enough, the real-time systems theory community briefly touched upon the topic of shared-memory multiprocessors, some 15 years ago, before putting the subject to sleep, after achieving results only in the negative (the so-called Dhall"s effect). The community interest was much revived when it appeared that hardware manufacturers would go multicore regardless of the possible disruption that such a move could cause to state-of-the-art real-time systems theory and industrial practice in the field. In this talk we will review how the state of the art in real-time systems theory has progressed over the last decade, and appreciate the alarming distance between what the abstract models understand, for the purpose of feasibility analysis in the time dimension, and what the actual system behaviour is at both hardware and software level. The talk with conclude with a short summary of the prospects opened, in the regard of those dire challenges, by very innovative research projects recently conducted by teams of international experts under the lead of BSC.

BIOGRAPHY
http://www.math.unipd.it/~tullio/

Conference information