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

Deep learning models can take weeks to train on a single GPU-equippedmachine, necessitating scaling out DL training to a GPU-cluster. However,current distributed DL implementations can scale poorly due to substantialparameter synchronization over the network, because the high throughput of GPUsallows more data batches to be processed per unit time than CPUs, leading tomore frequent network synchronization. We present Poseidon, an efficientcommunication architecture for distributed DL on GPUs. Poseidon exploits thelayered model structures in DL programs to overlap communication andcomputation, reducing bursty network communication. Moreover, Poseidon uses ahybrid communication scheme that optimizes the number of bytes required tosynchronize each layer, according to layer properties and the number ofmachines. We show that Poseidon is applicable to different DL frameworks byplugging Poseidon into Caffe and TensorFlow. We show that Poseidon enablesCaffe and TensorFlow to achieve 15.5x speed-up on 16 single-GPU machines, evenwith limited bandwidth (10GbE) and the challenging VGG19-22K network for imageclassification. Moreover, Poseidon-enabled TensorFlow achieves 31.5x speed-upwith 32 single-GPU machines on Inception-V3, a 50% improvement over theopen-source TensorFlow (20x speed-up).

Poseidon: An Efficient Communication Architecture for Distributed Deep Learning on GPU Clusters