Evaluating Performance Overheads of HPX/Kokkos Applications in Singularity Containers on Supercomputer Fugaku and a Hybrid CPU-GPU Cluster
Containerization can improve reproducibility of HPC applications, but may introduce performance overheads. This study evaluates the performance impact of running the HPX/Kokkos-based astrophysics application Octo-Tiger in Singularity containers on a homogeneous CPU-based supercomputer (Fugaku) and a heterogeneous CPU-GPU cluster (DeepBayou).
Scaling the AlphaFold Training to 2080 GPUs and Reducing Initial Training Time to 10 Hours
A systematic training method called ScaleFold that incorporates optimizations to address the key factors preventing the AlphaFold training from scaling to more compute resources, enabling it to be completed in 10 hours.
Identifying and Mitigating Slow Nodes in a Large Supercomputer Cluster Using Machine Learning and Proxy Applications
Identifying and mitigating the impact of slow-performing nodes in a large supercomputer cluster through the use of machine learning, proxy applications, and scheduling prioritization.
Wilkins: An Efficient and User-Friendly In Situ Workflow System for HPC
Wilkins is an in situ workflow system designed to provide scalable and efficient execution of diverse scientific tasks without requiring any modifications to user task codes.
Comprehensive Analysis of GPU Acceleration for Computational Fluid Dynamics on HPC Systems: Evaluating Speed, Power, and Cost Implications
This study provides a comprehensive analysis of GPU acceleration for computational fluid dynamics (CFD) simulations on HPC systems, evaluating the impact on simulation speed, power consumption, and cost.
Transforming Matrix Operations into Efficient Graph Processing for High-Performance Scientific Computing
Matrix operations can be transformed into equivalent graph representations, enabling domain experts to implement various types of matrix computations using a unified graph programming interface. This graph engine-based scientific computing paradigm achieves performance comparable to the best-performing implementations while greatly simplifying the development of scientific computations on large-scale platforms.
- 3d-モデル生成のための相互作用型-lmm-エージェントシステム
- 자연어와-이미지-3d-모델을-결합한-다중-모달-입력-idea-으로부터-효율적으로-3d-모델을-생성하는-idea-2-3d-프레임워크
- a-generalized-ziv-zakai-lower-bound-on-the-minimum-mean-squared-error
- idee-2-3d-automatisierte-erstellung-detaillierter-3d-modelle-aus-komplexen-multimodalen-eingaben
- 一般化されたziv-zakai下界に関する包括的な研究
- 추정-문제에-대한-포괄적인-ziv-zakai-하한-연구
- fine-grained-complexity-of-subgraph-isomorphism-characterizing-patterns-with-subquadratic-algorithms
- eine-umfassende-studie-zu-ziv-zakai-untergrenzschranken-für-den-mmse
© 2024 by Linnk AI