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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Recently, Graph Neural Networks (GNNs) have become state-of-the-artalgorithms for analyzing non-euclidean graph data. However, to realizeefficient GNN training is challenging, especially on large graphs. The reasonsare many-folded: 1) GNN training incurs a substantial memory footprint.Full-batch training on large graphs even requires hundreds to thousands ofgigabytes of memory. 2) GNN training involves both memory-intensive andcomputation-intensive operations, challenging current CPU/GPU platforms. 3) Theirregularity of graphs can result in severe resource under-utilization andload-imbalance problems.  This paper presents a GNNear accelerator to tackle these challenges. GNNearadopts a DIMM-based memory system to provide sufficient memory capacity. Tomatch the heterogeneous nature of GNN training, we offload the memory-intensiveReduce operations to in-DIMM Near-Memory-Engines (NMEs), making full use of thehigh aggregated local bandwidth. We adopt a Centralized-Acceleration-Engine(CAE) to process the computation-intensive Update operations. We furtherpropose several optimization strategies to deal with the irregularity of inputgraphs and improve GNNear's performance. Comprehensive evaluations on 16 GNNtraining tasks demonstrate that GNNear achieves 30.8$\times$/2.5$\times$geomean speedup and 79.6$\times$/7.3$\times$(geomean) higher energy efficiencycompared to Xeon E5-2698-v4 CPU and NVIDIA V100 GPU.

GNNear: Accelerating Full-Batch Training of Graph Neural Networks with Near-Memory Processing