BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.adass.org//adass2026//speaker//38JAVQ
BEGIN:VTIMEZONE
TZID:AWST
BEGIN:STANDARD
DTSTART:20000101T000000
RRULE:FREQ=YEARLY;BYMONTH=1;UNTIL=20051231T160000Z
TZNAME:AWST
TZOFFSETFROM:+0800
TZOFFSETTO:+0800
END:STANDARD
BEGIN:STANDARD
DTSTART:20070325T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:AWST
TZOFFSETFROM:+0900
TZOFFSETTO:+0800
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20071028T030000
RRULE:FREQ=YEARLY;BYDAY=4SU;BYMONTH=10;UNTIL=20081025T190000Z
TZNAME:AWDT
TZOFFSETFROM:+0800
TZOFFSETTO:+0900
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-adass2026-CNYTGX@pretalx.adass.org
DTSTART;TZID=AWST:20261104T111500
DTEND;TZID=AWST:20261104T113000
DESCRIPTION:## Galactic Plane Imaging and Polarimetry from MeerKAT Data on 
 High-Performance and Memory-Based Computing Resources\n\n**Authors:**  \nL
 . Haupt<sup>1</sup>\, A. Basu<sup>1</sup>\, S. Borra<sup>1</sup>\, E. Buch
 holz<sup>1</sup>\, H. Heßling<sup>1</sup>\, Y. K. Ma<sup>2</sup>\n\n**Aff
 iliation:**  \n<sup>1</sup> German Center for Astrophysics (DZA)\, Görlit
 z\, Germany\n<sup>2</sup> Max Planck Institute for Radio Astronomy (MPIfR)
 \, Bonn\, Germany\n\n**Date:**  \nJuly 28\, 2026\n\n---\n\n## Abstract\n\n
 The MPIfR–MeerKAT Galactic Plane Survey (MMGPS) produces large multidime
 nsional imaging datasets\, including full-polarisation image cubes and Far
 aday cubes. Processing these datasets requires high-capacity storage toget
 her with high-performance computing (HPC) resources\, as image calibration
  and reconstruction demand exceptionally large amounts of memory. Unfortun
 ately\, processing the complete MMGPS workflow currently takes several wee
 ks.\n\nAt the German Center for Astrophysics (DZA)\, a multidisciplinary t
 eam with expertise in astrophysics\, computer science\, and mathematics pe
 rformed a detailed analysis of the MMGPS processing pipeline with respect 
 to memory\, storage\, and computational requirements. The study revealed s
 ignificant I/O bottlenecks that limit the overall processing performance. 
 Our objective is to optimise resource allocation by mapping each pipeline 
 stage to the hardware architecture best suited for its requirements\, ther
 eby significantly improving efficiency without modifying the underlying sc
 ientific software.\n\nMemory-Based Computing (MBC) plays a central role in
  this approach. Unlike conventional HPC systems\, which are primarily proc
 essor-centric\, MBC is designed around a large shared-memory architecture.
  The DZA operates an MBC prototype equipped with 48 TB of shared main memo
 ry\, providing an ideal platform for memory-intensive radio astronomy appl
 ications.\n\nBy combining HPC and MBC resources\, we successfully reconstr
 ucted a complete three-dimensional MeerKAT image cube. Furthermore\, the l
 arge memory capacity of the MBC system enabled the generation of a Faraday
  cube and its direct visualisation as a movie. Overall\, the optimised map
 ping of pipeline stages to the appropriate hardware improved processing pe
 rformance by approximately one order of magnitude. In addition to these ha
 rdware-related improvements\, we present first considerations for potentia
 l software-level optimisations.\n\nFuture work will focus on identifying r
 emaining bottlenecks and evaluating parallelisation strategies. The MMGPS 
 pipeline serves as an ideal prototype for preparing data-processing workfl
 ows for the substantially larger datasets expected from the Square Kilomet
 re Array Observatory (SKAO).
DTSTAMP:20261001T101636Z
LOCATION:Banquet Hall
SUMMARY:Galactic Plane Imaging and Polarimetry from MeerKAT data on High-pe
 rformance and Memory-based computing resources - Elsa Buchholz (DZA)\, Lar
 s Haupt
URL:https://pretalx.adass.org/adass2026/talk/CNYTGX/
END:VEVENT
END:VCALENDAR