Scalable Node Embedding Algorithms Using Distributed Sparse Matrix Operations

We introduce a distributed memory parallel algorithm for force-directed node embedding that places vertices of a graph into a low-dimensional vector space based on the interplay of attraction among neighboring vertices and repulsion among distant vertices. We develop our algorithms using two sparse...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2024 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW) s. 1199 - 1201
Hlavní autoři: Ranawaka, Isuru, Azad, Ariful
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 27.05.2024
Témata:
Conferences
Data transfer
Distributed processing
Embedding
Parallel algorithms
SDDMM
Sparse matrices
Sparse Matrix Computation
SpMM
Supercomputers
Vectors
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:We introduce a distributed memory parallel algorithm for force-directed node embedding that places vertices of a graph into a low-dimensional vector space based on the interplay of attraction among neighboring vertices and repulsion among distant vertices. We develop our algorithms using two sparse matrix operations, SDDMM and SpMM. We propose a configurable pull-push-based communication strategy that optimizes memory usage and data transfers based on computing resources and asynchronous MPI communication to overlap communication and computation. Our algorithm scales up to 256 nodes on distributed supercomputers by surpassing the performance of state-of-the-art algorithms.
DOI:10.1109/IPDPSW63119.2024.00205