Parallel Computing in the era of the Cloud and Heterogeneous Computing

Ημερομηνία-χώρος: Παρασκευή 14 Δεκεμβρίου, 3-5μμ, Κτίριο Β (Αίθουσα Β4)

Ομιλητής: Rizos Sakellariou, Reader at the University of Manchester
Τίτλος: “Parallel Computing in the era of the Cloud and Heterogeneous Computing”

Περίληψη: Parallel computing has traditionally been the main approach used to speed up the performance of compute-intensive applications. However, as the complexity and the costs associated with modern execution platforms and applications grow, speeding up performance through parallelism cannot be viewed as a single objective to achieve at any cost. Instead, with such execution platforms consuming large amounts of energy, one needs to assess improvements in execution time against other types of cost. Cloud computing platforms, which are often used to execute compute- as well as data-intensive applications, typically follow a resource-on-demand paradigm, where users pay for what resources they need. Their increasing complexity is partly masked by having users pay sometimes for more than they need. In this respect, the talk will motivate the need to reflect upon and address efficiently the various trade-offs related to the concurrent use of multiple resources offered by Cloud providers as well as the characteristics of applications. Algorithms, techniques, experimental results and ongoing work will be presented in relation to the execution of compute- and data-intensive applications on the Cloud in such a way that different trade-offs such as execution time, energy consumption, monetary cost, etc are taken into account. The talk will conclude with examples motivating this way of thinking and research further.

Σχετικά με τον ομιλητή: Rizos Sakellariou obtained his PhD from the University of Manchester in 1997 for a thesis on compile-time parallel loop partitioning and scheduling. Since then, he has held posts with Rice University and the University of Cyprus and for the last 18 years with the University of Manchester where he is currently a Reader, leading a laboratory carrying our research in High-Performance, Parallel and Distributed systems, which over the last ten years has hosted more than 30 doctoral students, postdoctoral researchers and long-term visitors. He has carried out research on a number of topics related to parallel and distributed computing, has published more than 150 papers, has participated in more than 20 funded projects and been involved with the organization of over 150 conferences as committee member. Further information about his published work can be found from  https://scholar.google.com/citations?user=5LPRRM8AAAAJ http://www.cs.manchester.ac.uk/about-us/staff/profile/?ea=rizos

Μπορεί να δημοσιευθούν φωτογραφίες στη σελίδα του Τμήματος και στο Facebook.