Moris Behnam

Predictability and dependability in parallel architectures
Mälardalen University Sweden
15, May, 2018 15:30-16:30 (1 hour)
I 025, FEUP, Porto, Portugal

Many of today industrial software applications e.g., trains, autonomous utility vehicles, aviation, smart grid power management etc., require high performance, predictability and dependability guarantees. While the state of the art COTS hardware systems provide high performance through parallelization, dependability and predictability are still challenging. In this talk I will present some results, and ongoing activates within the "DPAC Dependable Platforms for Autonomous systems and Control" project. Multicore and heterogeneous system (including GPU and FPGA) architectures are considered. We investigate how heavy parallel applications use low-level resources (e.g., machine learning, video feature detection algorithms) and try to optimize the scheduling and allocation accordingly. In addition, we investigate how to provide different levels of isolation between different applications sharing the system resources.

Dr. Moris Behnam is a Senior Lecturer at MDH. He has awarded a B.Eng., and M.Sc. in Computer and Control Engineering at the University of Technology, Iraq, and also MS.c., Licentiate, and PhD in Computer Science and Engineering at MDH, Sweden, in 1995, 1998, 2005, 2008 and 2010 respectively. Moris has been a visiting researcher at Wayne State University, USA and he has been a Postdoctoral Researcher at University of Porto.
His main research interest is on resource virtualization for industrial distributed real-time in addition to real-time scheduling, synchronization protocols, multicore/multiprocessor systems, distributed embedded real-time sysems, indutrial cloud computing and internet of things. Regarding his research record, Moris has published more than 160 journal/conference/workshop papers.

At CISTER's Facebook page / At CISTER's Instagram page