Spatially Parallel Convolutions

By Peter H. Jin, Boris Ginsburg, Kurt Keutzer. 2018

10 citations recorded by the source corpus.

Subfields: Model Architectures

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Model Architectures

1. Noam Shazeer
2. Kurt Keutzer
3. Ashish Vaswani
4. Niki Parmar
5. A. Gholami
6. Jianfeng Gao
7. Quoc V. Le
8. Cliff Young

Citations in the corpus

• Mesh-TensorFlow: Deep Learning for Supercomputers (2018, background): “For example, convolutions on spatiallypartitioned tensors will require the communication of "halo" regions, as described in [21].”
• Integrated Model, Batch, and Domain Parallelism in Training Neural Networks (2017, methodology): “A third possibility for parallelization is domain parallel [12], where one can decompose the input activation map as shown in Fig.”