Scalable Algorithms Using Sparse Storage for Parallel Spectral Clustering on GPU - CentraleSupélec Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Scalable Algorithms Using Sparse Storage for Parallel Spectral Clustering on GPU

Résumé

Spectral clustering has many fundamental advantages over k-means, but has high computational complexity (O(n**3)) and memory requirement (O(n**2)), making it prohibitively expensive for large datasets. In this paper we present our solution on GPU to address the scalability challenge of spectral clustering. First, we propose optimized algorithms for constructing similarity matrix directly in CSR sparse format on the GPU. Next, we leverage the spectral graph partitioning API of the GPU-accelerated nvGRAPH library for remaining computations especially for eigenvector extraction. Finally, experiments on synthetic and real-world large datasets demonstrate the high performance and scalability of our GPU implementation for spectral clustering.
Fichier principal
Vignette du fichier
Scalable Algorithms Using Sparse Storage for Parallel Spectral Clustering on GPU.pdf (1.84 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04138695 , version 1 (23-06-2023)

Identifiants

Citer

Guanlin He, Stephane Vialle, Nicolas Sylvestre, Marc Baboulin. Scalable Algorithms Using Sparse Storage for Parallel Spectral Clustering on GPU. IFIP International Conference on Network and Parallel Computing (NPC 2021), Christophe Cérin; Depei Qian; Jean-Luc Gaudiot; Guangming Tan; Stéphane Zuckerman, Nov 2021, Paris, France. pp.40-52, ⟨10.1007/978-3-030-93571-9_4⟩. ⟨hal-04138695⟩
3 Consultations
5 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More