BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//TYPO3/NONSGML Calendarize//EN
BEGIN:VEVENT
UID:calendarize-verlaessliche-methoden-auf-grafikprozessoren-im-kontext-en
 ergietechnischer-systeme
DTSTAMP:20260908T102608Z
DTSTART:20260918T140000Z
SUMMARY:Reliable methods on graphics processing units in the context of en
 ergy systems
DESCRIPTION:On Friday 18 September 2026 at 2.30 pm\, \nLorenz Gillner\, Un
 iversity of Wismar\,\nwill give a lecture as part of his planned PhD thesi
 s\, entitled\nReliable Methods on Graphics Processing Units in the Context
  of Energy SystemsThe lecture will take place in a hybrid format.\nIQON\, 
 Industriestr. 11\, Room 1-Meet  and https://us02web.zoom.us/j/2972318676?p
 wd=dmVIZjBOM2ltYjRNYkduS2JXYUJaUT09\nThe lecture will be delivered in Germ
 an.\nAbstract:Uncertainties are ubiquitous in scientific computing. From m
 easurement inaccuracies andmodel uncertainty to discretisation errors aris
 ing from the use of computers with finiteprecision\, uncertainty affects t
 he reliability of computer-aided calculations. Greaterreliability can be a
 chieved by treating uncertainties as an integral part of the system.Set-th
 eoretic techniques\, such as interval methods\, enable calculations involv
 ing quantities subject to uncertaintythat are nevertheless verified. At th
 e same time\, growing demands on computer systemsare driving a transformat
 ion of traditional computer architecture\, thereby enabling new approaches
  toproblem-solving. Graphics processing units (GPUs)\, for example\, have 
 long since ceased to be used exclusively in thefield of computer graphics.
  Thanks to their ability to provide hardware-based accelerationof parallel
 isable operations\, they have evolved into an integral part of modern hete
 rogeneoushigh-performance computing systems. However\, computations on co-
 processors such asGPUs remain largely unverified. Established software and
  algorithms must therefore be reconsidered from a newperspective in order 
 to achieve a computational advantage—both in terms of theproblem complex
 ity and the efficiency of the solution. The aim of this PhD project is the
  applicationof GPU-accelerated interval methods for the reliable simulatio
 n and optimisation of dynamicsystems. This talk focuses on the further dev
 elopment of conventional interval software forheterogeneous computing syst
 ems and demonstrates how the use of the GPU can counteract the practical h
 urdles associated withinterval methods. The methods developed are tested f
 or their practical suitability using the example ofparameter estimation fo
 r a battery cell.\nSupervisors: Prof. Dr.-Ing. habil. Andreas Rauh (UOL) a
 nd Prof. Dr. rer. nat. habil. Ekaterina Auer (Wismar University of Applied
  Sciences)
X-ALT-DESC;FMTTYPE=text/html:<div style="margin-left:0px\;"><p class="x_De
 fault">On <strong>Friday 18 September 2026 at 2.30 pm\, </strong></p>\n<p 
 class="x_Default"><strong>Lorenz Gillner\, University of Wismar\,</strong>
 </p>\n<p class="x_Default">will give a lecture as part of his planned PhD 
 thesis\, entitled</p>\n<p class="x_Default"><strong>Reliable Methods on Gr
 aphics Processing Units in the Context of Energy Systems</strong><br />The
  lecture will take place in a hybrid format.</p>\n<p class="x_Default"><st
 rong>IQON\, Industriestr. 11\, Room 1-Meet  and </strong><a class="OWAAuto
 Link" href="https://us02web.zoom.us/j/2972318676?pwd=dmVIZjBOM2ltYjRNYkduS
 2JXYUJaUT09"><strong>https://us02web.zoom.us/j/2972318676?pwd=dmVIZjBOM2lt
 YjRNYkduS2JXYUJaUT09</strong></a><strong><br /></strong></p>\n<p class="x_
 Default">The lecture will be delivered in German.</p>\n<p><strong>Abstract
 </strong>:<br />Uncertainties are ubiquitous in scientific computing. From
  measurement inaccuracies and<br />model uncertainty to discretisation err
 ors arising from the use of computers with finite<br />precision\, uncerta
 inty affects the reliability of computer-aided calculations. Greater<br />
 reliability can be achieved by treating uncertainties as an integral part 
 of the system.<br />Set-theoretic techniques\, such as interval methods\, 
 enable calculations involving quantities subject to uncertainty<br />that 
 are nevertheless verified. At the same time\, growing demands on computer 
 systems<br />are driving a transformation of traditional computer architec
 ture\, thereby enabling new approaches to<br />problem-solving. Graphics p
 rocessing units (GPUs)\, for example\, have long since ceased to be used e
 xclusively in the<br />field of computer graphics. Thanks to their ability
  to provide hardware-based acceleration<br />of parallelisable operations\
 , they have evolved into an integral part of modern heterogeneous<br />hig
 h-performance computing systems. However\, computations on co-processors s
 uch as<br />GPUs remain largely unverified. Established software and algor
 ithms must therefore be reconsidered from a new<br />perspective in order 
 to achieve a computational advantage—both in terms of the<br />problem c
 omplexity and the efficiency of the solution. The aim of this PhD project 
 is the application<br />of GPU-accelerated interval methods for the reliab
 le simulation and optimisation of dynamic<br />systems. This talk focuses 
 on the further development of conventional interval software for<br />hete
 rogeneous computing systems and demonstrates how the use of the GPU can co
 unteract the practical hurdles associated with<br />interval methods. The 
 methods developed are tested for their practical suitability using the exa
 mple of<br />parameter estimation for a battery cell.</p>\n<p><br /><stron
 g>Supervisors:</strong> Prof. Dr.-Ing. habil. Andreas Rauh (UOL) and Prof.
  Dr. rer. nat. habil. Ekaterina Auer (Wismar University of Applied Science
 s)</p></div>
LOCATION:
END:VEVENT
END:VCALENDAR