English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The increasing size of input graphs for graph neural networks (GNNs)highlights the demand for using multi-GPU platforms. However, existingmulti-GPU GNN systems optimize the computation and communication individuallybased on the conventional practice of scaling dense DNNs. For irregularlysparse and fine-grained GNN workloads, such solutions miss the opportunity tojointly schedule/optimize the computation and communication operations forhigh-performance delivery. To this end, we propose MGG, a novel system designto accelerate full-graph GNNs on multi-GPU platforms. The core of MGG is itsnovel dynamic software pipeline to facilitate fine-grainedcomputation-communication overlapping within a GPU kernel. Specifically, MGGintroduces GNN-tailored pipeline construction and GPU-aware pipeline mapping tofacilitate workload balancing and operation overlapping. MGG also incorporatesan intelligent runtime design with analytical modeling and optimizationheuristics to dynamically improve the execution performance. Extensiveevaluation reveals that MGG outperforms state-of-the-art full-graph GNN systemsacross various settings: on average 4.41X, 4.81X, and 10.83X faster than DGL,MGG-UVM, and ROC, respectively.

MGG: Accelerating Graph Neural Networks with Fine-grained intra-kernel Communication-Computation Pipelining on Multi-GPU Platforms