2026-09-28 10:27 |
Details des Eintrags - Ähnliche Datensätze
|
2026-09-28 10:22 |
Details des Eintrags - Ähnliche Datensätze
|
2026-09-24 08:38 |
Details des Eintrags - Ähnliche Datensätze
|
2026-09-21 11:31 |
[FZJ-2026-04493]
Master Thesis
Suresh, A. ; Saviankou, P. ; Plessl, C. ; et al
LLM in CubeGUI for HPC Performance Analysis
Masterarbeit, Universität Padeborn, 2026
High-performance computing (HPC) applications often suffer from performanceissues such as communication overhead, synchronization delays, load imbalance,serialization, etc., which are traditionally identified through manual and expert-driven analysis of profiling data. This work investigates the utility of large lan-guage models (LLMs) for analyzing performance problems in parallel programswith load imbalance analysis as a representative use case.To explore the use of Large Language Models (LLM) in parallel applications,structured performance data is extracted from CubeGUI (a GUI that displays per-formance data for parallel programs including MPI and OpenOpenMP applica-tions). [...]
Restricted: PDF;
Details des Eintrags - Ähnliche Datensätze
|
2026-09-16 11:08 |
Details des Eintrags - Ähnliche Datensätze
|
2026-09-16 10:14 |
Details des Eintrags - Ähnliche Datensätze
|
2026-09-11 09:20 |
Details des Eintrags - Ähnliche Datensätze
|
2026-09-10 13:23 |
[FZJ-2026-04392]
Journal Article
Zhou, H. ; Wang, S. ; Huo, M. ; et al
Learning when to use physical relationships in scientific machine learning
AI for Science 0, 0 pp. (2026)2026
Multiple properties of materials and energy devices are linked by known physical relationships, yet machine-learning models commonly predict each property independently from descriptors, here termed Direct prediction. This study tests whether these relationships remain useful when their inputs are themselves model predictions. [...]
Details des Eintrags - Ähnliche Datensätze
|
2026-09-10 13:16 |
[FZJ-2026-04391]
Journal Article
Zhou, H. ; Yu, S. ; Liaw, B. ; et al
Intrinsic Lattice Formation Strain Guides Screening of Sulfide Solid Electrolytes
npj Compuational Materials 0, 0 pp. (2026)2026
Efficient solid-electrolyte discovery requires a tractable, interpretable link between composition, structure and ion transport despite sparse datasets. Here, intrinsic lattice formation strain (ILFS) is introduced to provide a reduced, family-referenced structural coordinate defined by the relative volumetric deviation between an experimentally refined structure and a symmetry-limited virtual reference. [...]
Details des Eintrags - Ähnliche Datensätze
|
2026-09-10 13:12 |
Details des Eintrags - Ähnliche Datensätze
|
|
|