Pieter Noordhuis
Emeritus maintainer of PyTorch Distributed; contributor to the Gloo collective communications library
Pieter Noordhuis is listed among the emeritus maintainers of the Distributed component on PyTorch's persons-of-interest page. Noordhuis is a co-author of the VLDB 2020 paper on PyTorch's distributed data parallel module and of the 2017 paper on training ImageNet with large minibatches in about an hour. Noordhuis's commits to the Gloo collective communications library include one that created its libuv-based transport in 2019. No current organizational role is documented in the sources reviewed.1345
Work
PyTorch DistributedOpen-source code
Listed as an emeritus maintainer of PyTorch's Distributed module, and a co-author of "PyTorch Distributed: Experiences on Accelerating Data Parallel Training" (VLDB 2020), which describes the design of the distributed data parallel module. Noordhuis's GitHub profile pins the pytorch/pytorch repository.132
Gloo (external site: github.com)Open-source code
Author of commits to the BSD-licensed Gloo library; those listed from May 2019 to January 2020 include "Create libuv transport" and changes to the TCP transport.56
Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour (external site: arxiv.org)Published research
Co-author of the 2017 paper, which trains ResNet-50 on ImageNet with minibatches of up to 8,192 images using a learning-rate scaling rule and warmup.4
Other documented roles
Roles the sources document but that are not confirmed as current.
- Maintainer, Distributed (now listed as emeritus), PyTorch (open-source project)1
This profile covers public professional work only, from the sources below. It is not an endorsement or a ranking. To correct or remove it, use the correction form.