Conference Presentation (Other) FZJ-2026-05825

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png
Dense Numerical Linear Algebra under the Microscope: what applications really need

 ;

2026

Platform for Advanced Scientific Computing 2026, PASC26, BernBern, Switzerland, 29 Jun 2026 - 1 Jul 20262026-06-292026-07-01 ()

Abstract: We are currently witnessing a substantial prevalence of GPU accelerated architectures among the current and future exascale supercomputers. This paradigm shift is now more than ever highlighting the need for a renewed leadership in numerical algebra software. In the EU ecosystem, the introduction of JUPITER, the first exascale booster in Europe, emphasizes the existing deficiency in distributed dense Numerical Linear Algebra libraries (NLA) that can effectively harness its enormous computational power. As application software is preparing to exploit these massively accelerated platforms, there is a critical need for software solutions that can manage and optimize computations at this scale. The lack of a clear candidate for such NLA library amounts to an entry-point barrier to fully utilizing exascale and post-exascale capabilities for scientific research and industrial applications, where distributed numerical linear algebra is essential for solving complex problems.In order to address this challenge, a groups of international experts launched in 2025 the ExaNLA initiative. Through its committee ExaNLA plans to define, oversee, and coordinate the core activities required to design and develop a high-performance, scalable, and portable dense Numerical Linear Algebra (NLA) library optimized for modern distributed and accelerated computing environments. The committee supports the creation of working groups whose activities may include the study and design of existing and novel algorithms, the evaluation and integration of established and emerging programming models, the formulation of benchmarks and testing protocols, and the development of community-driven documentation and support strategies.


Contributing Institute(s):
  1. Jülich Supercomputing Center (JSC)
Research Program(s):
  1. 5111 - Domain-Specific Simulation & Data Life Cycle Labs (SDLs) and Research Groups (POF4-511) (POF4-511)
  2. Simulation and Data Laboratory Quantum Materials (SDLQM) (SDLQM) (SDLQM)

Appears in the scientific report 2026
Click to display QR Code for this record

The record appears in these collections:
Dokumenttypen > Präsentationen > Konferenzvorträge
Workflowsammlungen > Öffentliche Einträge
Workflowsammlungen > In Bearbeitung
Institutssammlungen > JSC
Online First

 Datensatz erzeugt am 2026-11-10, letzte Änderung am 2026-11-13


Restricted:
Volltext herunterladen PDF
Dieses Dokument bewerten:

Rate this document:
1
2
3
 
(Bisher nicht rezensiert)