<html><body><div style="color: rgb(0, 0, 0); font-family: arial,helvetica,sans-serif; font-size: 12pt;"><div><br><b>Sent: </b>Wednesday, January 21, 2015 3:07:53 PM<br><b>Subject: </b>Re: ODU CAS seminar- Tomorrow<br></div><div><br></div><div data-marker="__QUOTED_TEXT__"><div class="moz-cite-prefix">Hello All,<br> <br> Tomorrow's seminar will also have a video feed in the MCC Conference Room. <br> <br> Regards,<br> <br> Crystal <br><div class="moz-signature"><br><div class="moz-signature">-- <br> <img alt="" src="cid:part1.06040204.06050907@jlab.org" width="273" height="177" data-mce-src="cid:part1.06040204.06050907@jlab.org"></div></div>On 1/21/2015 11:46 AM, Crystal Baker wrote:<br></div><blockquote cite="mid:54BFD7E8.8000902@jlab.org"><br> Dear Colleagues: <br> <br> We will have our next CAS seminar this coming Thursday at 1:30 at ODU in the CAS building (see below). There will also be a video link for the talk at Jefferson Lab in the ARC building in room 728. Pizza and other refreshments will be available at ODU at 1 pm. The schedule for future talks is listed at the end of this message. If you want to be added to this email distribution list, please let me know. <br> <br> High Performance Computing with Application to Physics <br> Dr. Desh Ranjan & Dr. Mohammad Zubair <br> Department of Computer Science <br> Old Dominion University <br> <br> When: Thursday January 22, 2015; 1:30 pm (Pizza/Refreshments at 1 pm) <br> Where: CAS Building at ODU (1021 W. 47th St.) <br> Video: Jefferson Lab ARC 728 <br> <br> Abstract: The first part of the talk deals with addressing memory bottleneck in parallel computing to achieve high performance computing. In the second part of the talk we look at a memory-efficient algorithm and its implementation for solving multidimensional numerical integration on a cluster of compute nodes with multiple GPU devices per node. Multidimensional numerical integration is one of the most important and widely used computational tools in various fields of computational science such as simulation of coherent synchrotron radiation in charged particle beams via multidimensional space-time integration of retarded potentials. The effective use of shared memory is important for improving the performance on GPUs, because of the bandwidth limitation of the global memory. __________________________________________________________ <br> Future Speakers: <br> Feb. 26: Joe Grames <br> Mar. 26: Janardan Upadhyay <br> <br> <br> --------------------------------------- <br> Gail Dodge <br> Department of Physics <br> Old Dominion University <br> Norfolk, VA 23529 <br> <br></blockquote><br><br></div></div></body></html>