BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//ZContent.net//ZapCalLib 1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
DTSTART;TZID=Atlantic/Canary:20120329T000000
DTEND;TZID=Atlantic/Canary:20120329T010000
UID:iactalks-375
X-WR-CALNAME: IAC Talks: Open Astronomy Seminars
X-ORIGINAL-URL: /iactalks/Talks/view/375
CREATED:2012-03-29T00:00:00+01:00
X-WR-CALDESC: IAC Talks upcomming talks
SUMMARY:The Future of High Performance Computing for Astrophysicists
DESCRIPTION:The Future of High Performance Computing for Astrophysicists\nD
 r. Lars Koesterke\n\nWith the advent of GPU accelerators the landscape of 
 High Performance  Computing has started to change rapidly. While this is i
 n principle good  news, the increased compute power comes with a steep pri
 ce tag in that  new languages (CUDA, OpenCL) must be used. Recently Intel 
 has announced  their own coprocessor Many Integrated Cores (MIC) technolog
 y which will  deliver competitive performance but will be programmed throu
 gh familiar  languages (Fortran, C/C++ and OpenMP). In my talk I will intr
 oduce Intel's MIC architecture and will discuss  the ongoing efforts at th
 e Texas Advanced Computing Center to build a 10  PetaFlop cluster with MIC
  coprocessors in early 2013. Coprocessors  (MIC) and accelerators (GPU) ar
 e here to stay and the changing hardware  will spur considerable changes i
 n general software design. Astrophysics  codes of all varieties (for examp
 le highly parallel simulations,  data-intensive software pipelines for lar
 ge surveys, and even data  reduction software on desktops) will have to ad
 apt to the new  environment. I will discuss software design, performance c
 onsiderations,  and optimizations in general and specifically with respect
  to the MIC  technology.  In the second part of my talk I will introduce t
 he software package  ASSET (Advanced Spectral Synthesis 3D Tool) that allo
 ws for the fast and  efficient calculation of spectra from 3D hydrodynamic
 al models and will  highlight recent projects that have employed high-reso
 lution (&gt;  1,000,000), wide-range (1000's of Angstroem) synthetic spect
 ra derived  from 3D radiation transfer.
END:VEVENT
END:VCALENDAR
