BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20250822T115808Z
LOCATION:Room 5.0B15 & 16
DTSTART;TZID=Europe/Stockholm:20250617T150000
DTEND;TZID=Europe/Stockholm:20250617T170000
UID:submissions.pasc-conference.org_PASC25_sess135@linklings.com
SUMMARY:MS4B - Julia for HPC: Reproducible High-Performance Computing
DESCRIPTION:The fourth instalment of the “Julia for HPC” PASC minisymposiu
 m focuses on reproducibility, a cornerstone of scientific research and a k
 ey component of High-Performance Computing (HPC). As hardware and software
  evolve rapidly, reproducibility of both scientific results and applicatio
 n performance becomes increasingly complex. This challenge is particularly
  pronounced in HPC, where software packages are often tailored to specific
  hardware architectures for optimal performance. Addressing reproducibilit
 y in HPC therefore necessitates the careful development of portable librar
 ies and applications, portable packaging, and a package management system 
 that controls versions of all kinds of software dependencies. This minisym
 posium highlights the Julia programming language and its environment, whic
 h tackle these interconnected challenges in a holistic manner. Tightly int
 egrated tools, such as the package manager and artifact builder, coordinat
 e packaging and environment management, offering a consistent approach. Ex
 pert speakers will provide insights into the current state of reproducibil
 ity in Julia, highlighting strengths and remaining hurdles. The talks will
  be presented from the perspectives of both tooling developers and domain 
 scientists. This minisymposium is aimed at Julia users eager to deepen the
 ir understanding of the existing reproducibility toolchain, as well as non
 -Julia users curious about how Julia’s reproducibility solutions might be 
 adapted to other software ecosystems.\n\nReproducible Heterogeneous Comput
 ing with the Julia Language\n\nI will talk about IPUToolkit.jl, a package 
 for running Julia code on the Intelligence Processing Unit (IPU), a massiv
 ely parallel accelerator developed by Graphcore and powered by almost 1500
  cores. I will show how Julia enables a high-degree of code reuse on this 
 specialised hardware, using advance...\n\n\nMose Giordano (University Coll
 ege London)\n---------------------\nParallelizing GaPSE.jl with KernelAbst
 raction.jl: A Real-World Example of Reproducibility in Julia\n\nJulia is g
 aining traction in scientific computing, and at the Leibniz Supercomputing
  Centre (LRZ), we are exploring its potential on our high-performance comp
 uting (HPC) system, particularly on our Intel Ponte Vecchio GPUs of the Su
 perMUC-NG Phase 2 supercomputer. The Julia package KernelAbstraction...\n\
 n\nMatteo Foglieni (Leibniz Supercomputing Centre)\n---------------------\
 nGenerating Architecture-Agnostic Performance Tests from Functional Unit T
 ests Using Classical Performance Models\n\nPerfTest.jl is a Julia package 
 conceived from the idea of bridging the gap between unit testing and  arch
 itecture-agnostic performance testing. It brings a set of features that al
 low the user to set up performance regression unit tests from functional u
 nit tests with minimal effort. An emphasis is m...\n\n\nDaniel Sergio Vega
  Rodriguez (Università della Svizzera italiana), Samuel Omlin (ETH Zurich 
 / CSCS), and Dimosthenis Pasadakis and Olaf Schenk (Università della Svizz
 era italiana)\n---------------------\nA GPU-Accelerated Unified API for Si
 ngular Values Enabling Reproducability Across Architectures and Data Types
 \n\nWe present a portable, GPU-accelerated implementation of a QR-based Si
 ngular Value algorithm in Julia, that allows code reproducibility across s
 everal different GPU vendors. Singular Value Decomposition (SVD) is a fund
 amental numerical tool in scientific computing and machine learning, provi
 ding opt...\n\n\nEvelyne Ringoot, Rabab Alomairy, Valentin Churavy, and Al
 an Edelman (Massachusetts Institute of Technology)\n\nDomain: Chemistry an
 d Materials, Engineering, Life Sciences, Physics, Computational Methods an
 d Applied Mathematics\n\nSession Chair: Samuel Omlin (ETH Zurich / CSCS)
END:VEVENT
END:VCALENDAR