ASPHD_CSC Archives

PhD Student

ASPHD_CSC@LISTSERV.GSU.EDU

Options: Use Forum View

Use Monospaced Font
Show Text Part by Default
Condense Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Sender:
PhD Student <[log in to unmask]>
Date:
Mon, 12 Apr 2010 14:34:20 -0400
Reply-To:
PhD Student <[log in to unmask]>
Subject:
MIME-Version:
1.0
Content-Transfer-Encoding:
8bit
Content-Type:
text/plain; charset=windows-1252; format=flowed
From:
Tammie Dudley <[log in to unmask]>
Comments:
Parts/Attachments:
text/plain (22 lines)
Thursday, April 15 – Departmental Colloquium
1:00 p.m.–3:00 p.m., Department Conference Room
"Programming with Parallel Tasks"
Dr. Gudula Rünger, Technische Universität Chemnitz

-----------------------------------------------------------
Programming with Parallel Tasks

Dr. Gudula Rünger
Fakultät für Informatik
Technische Universität Chemnitz
Chemnitz, Germany

Recent and future parallel clusters and supercomputers use SMPs and multicore processors as basic nodes, providing a huge amount of parallel resources. These systems often have hierarchically structured interconnection networks combining computing resources at different levels, starting with the interconnect within multicore processors up to the interconnection network combining nodes of the cluster or supercomputer. The challenge for the programmer is that these computing resources should be utilized efficiently by exploiting the available degree of parallelism of the application programs and by structuring the application in a way which is sensitive to the heterogeneous interconnect. In the talk, we pursue a parallel programming method using parallel tasks to structure parallel implementations. A parallel task can be executed by multiple processors or cores and, for each activation of a parallel task, the actual number of executing cores can be adapted to the specific execution situation. The proposed programming approach distinguishes between the specification of parallel tasks with their dependencies and the scheduling and mapping to execution units. An experimental evaluation shows that the scalability of parallel applications can be significantly improved by using suitable scheduling and mapping techniques.

About the Speaker: Prof. Gudula Rünger received her Diploma and her Ph.D. in mathematics from the University Cologne, Germany, in 1985 and 1989, respectively. From 1989 to 1997, she was with the University des Saarlandes, Germany, where she received her habilitation in Computer Science in 1996. In August 1997 she joined the University of Leipzig, Germany, where she was a professor in the area of parallel computing. Since 2000, she is a full professor at Chemnitz University of Technology, Germany. Her research interests are mainly focused on parallel programming and scientific computing.


thanks,

yanqing zhang

ATOM RSS1 RSS2