[hpc-announce] Call for Papers: PDML19 Workshop (in conjunction with ICPP2019 at Kyoto)

Kento Sato kento.sato at riken.jp
Sun Mar 10 20:09:51 CDT 2019

We apologize if you receive multiple copies of the CFP.

                                             CALL FOR PAPERS

The 1st Workshop on Parallel and Distributed Machine Learning  2019 (PDML19)
                                 Kyoto, Japan on August 5th, 2019

Held in conjunction with The 48th International Conference on Parallel Processing

- Paper Submission: May 13th, 2019 (AoE)
- Author Notification: May 31st, 2019
- Camera-Ready Copy: June 7th, 2019 
- Workshop Date: August 5th, 2019

Parallel and distributed computing has been making tremendous impacts on the recent advancement of data-oriented machine learning such as deep learning. Accelerating ML workloads with HPC systems can present opportunities to enable more complicated machine learning. However, significant challenges remain to be addressed due to limited computation power against the huge volume of datasets. In this workshop, we bring together researchers in the field of machine learning and facilitate discussions for their experiences, new ideas and the latest trends to leverage HPC for ML, ML for HPC and ML applications in HPC.

We welcome all audience who are interested in ML and HPC. Especially, we target researchers and practitioners that are actively working on applying parallel and distributed computing to machine learning. Specifically, the topics will include but not limited to:
- Algorithmic techniques to improve performance and efficiency of parallel applications of machine learning
- ML-based techniques to improve system and application efficiency of HPC environments
- Development of algorithms, models and solvers for parallel and distributed applications using machine/deep learning
- All aspects of parallel processing hardware including the optimization and evaluation of processors and networks for machine/deep learning
- Techniques for performance measurement, performance modeling and performance tools in machine/deep learning
- Techniques to support parallel programming, system software, runtime system and other low-level software research and development for machine/deep learning
- We also encourage submissions in emerging fields that may not fit into these categories to have more diversity of the topics. If authors are in doubt, we willing to have any questions from the authors.

We encourage the submission of both full and short papers containing high-quality research describing original and unpublished work. Short papers are intended to provide opportunities to present and discuss preliminary research results on emerging topics. Submissions should be in PDF format in U.S. letter size paper using the ACM conference style. Full and short papers are a maximum of eight (8) and four (4) double-column pages, respectively. Page limits include all figures, tables and appendices; only references do not count against the page limit. Submissions will be judged based on relevance, significance, originality, correctness and clarification. Reviews are not double-blind. All accepted papers are planned to be published by ACM, and included in ACM digital library if presented at the conference.

The paper submission online system is open: https://easychair.org/conferences/?conf=pdml19

- Naoya Maruyama, Lawrence Livermore National Laboratory
- Rio Yokota, Tokyo Institute of Technology
- Kento Sato, RIKEN Center for Computational Science

- Tal Ben-Nun, ETH Zurich
- Keisuke Fukuda, Preferred Networks
- Masaaki Kondo, University of Tokyo/RIKEN Center for Computational Science
- Naoya Maruyama, Lawrence Livermore National Laboratory
- Akira Naruse, NVIDIA
- Kento Sato, RIKEN Center for Computational Science
- Koichi Shirahata, Fujitsu Laboratories
- Mohamed Wahib, National Institute for Advanced Industrial Science and Technology
- Rio Yokota, Tokyo Institute of Technology


More information about the hpc-announce mailing list