BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//juliacon-2026//speaker//PLZXRD
BEGIN:VTIMEZONE
TZID:CET
BEGIN:STANDARD
DTSTART:20001029T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-juliacon-2026-ZLZ8FJ@pretalx.com
DTSTART;TZID=CET:20260813T161500
DTEND;TZID=CET:20260813T163000
DESCRIPTION:The Density-Functional ToolKit (DFTK) is a Julia package provid
 ing routines to compute\nthe electronic structure of a bulk material and r
 elated properties\,\nusing plane-wave density functional theory (DFT).\nMa
 ny material properties of interest can be expressed as derivatives of simu
 lation outputs\nwrt. input parameters\, and typically only specific combin
 ations are implemented by DFT codes\,\nas a result of great programming ef
 fort to hand-implement all the required derivative terms.\nIn DFTK however
 \, derivatives of **any** output quantity wrt. **any** input parameter can
  be computed\,\nusing algorithmic differentiation (AD) combined with densi
 ty-functional perturbation theory (DFPT).\nThis results in a general AD-DF
 PT framework [1] that can only be used to compute both standard and novel 
 derivatives\,\nwith promising applications including gradient-based optimi
 zation and error propagation.\n\nIn the first part of this talk\, I will d
 iscuss the key ideas behind this implementation\,\nshowing how we offload 
 tedious derivative computations to the AD framework\,\nwhile keeping the n
 umerics under control thanks to the underlying DFPT solver.\nThe overall s
 trategy is quite general\, and should be applicable in other fields as wel
 l.\nIn the second part of this talk\, I will present new research directio
 ns enabled by AD-DFPT.\nIn particular\, I will focus on the propagation of
  model parameter uncertainty \nand estimated numerical errors all the way 
 to predicted physical quantities.
DTSTAMP:20260428T134259Z
LOCATION:Room 2
SUMMARY:Algorithmic differentiation and error control with DFTK - Bruno Plo
 umhans
URL:https://pretalx.com/juliacon-2026/talk/ZLZ8FJ/
END:VEVENT
END:VCALENDAR