The Parallelization Project - New designs of domain specific accelerators and their compilers
As state-of-the-art smartphones, tablet, laptops and servers adopt multicore processors with GPUs and custom accelerators, computer systems become increasingly parallel and heterogeneous. The parallelization project investigates domain specific programs such as deep learning and scientific applications to find hardware acceleration opportunities, and proposes new hardware acceleration architecture and their compilers.

People in the Parallelization project

Publications

Refereed International Conference Publications

Context-Aware Memory Profiling for Speculative Parallelism [abstract] (IEEE Xplore, PDF)
Changsu Kim, Juhyun Kim, Juwon Kang, Jae W. Lee, and Hanjun Kim
Proceedings of the 24th IEEE International Conference on High Performance Computing, Data, and Analytics (HiPC), December 2017.

GPUpd: A Fast and Scalable Multi-GPU Architecture Using Cooperative Projection and Distribution [abstract] (ACM DL)
Youngsok Kim, Jae-Eon Jo, Hanhwi Jang, Minsoo Rhu, Hanjun Kim, and Jangwoo Kim
Proceedings of the 50th Annual IEEE/ACM International Symposium on Microarchitecture (MICRO), October 2017.

Practical Automatic Loop Specialization [abstract] (ACM DL, PDF)
Taewook Oh, Hanjun Kim, Nick P. Johnson, Jae W. Lee, and David I. August
Proceedings of the Eighteenth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), March 2013.

Speculative Separation for Privatization and Reductions [abstract] (ACM DL, PDF)
Nick P. Johnson, Hanjun Kim, Prakash Prabhu, Ayal Zaks, and David I. August
Proceedings of the 33rd ACM SIGPLAN Conference on Programming Language Design and Implementation (PLDI), June 2012.

Automatic Speculative DOALL for Clusters [abstract] (ACM DL, PDF)
Hanjun Kim, Nick P. Johnson, Jae W. Lee, Scott A. Mahlke, and David I. August
Proceedings of the 2012 International Symposium on Code Generation and Optimization (CGO), March 2012.

A Survey of the Practice of Computational Science [abstract] (ACM DL, PDF)
Prakash Prabhu, Thomas B. Jablin, Arun Raman, Yun Zhang, Jialu Huang, Hanjun Kim, Nick P. Johnson, Feng Liu, Soumyadeep Ghosh, Stephen Beard, Taewook Oh, Matthew Zoufaly, David Walker, and David I. August
Proceedings of the 24th ACM/IEEE Conference on High Performance Computing, Networking, Storage and Analysis (SC), November 2011.

Parallelism Orchestration using DoPE: the Degree of Parallelism Executive [abstract] (ACM DL, PDF)
Arun Raman, Hanjun Kim, Taewook Oh, Jae W. Lee, and David I. August
Proceedings of the 32nd ACM SIGPLAN Conference on Programming Language Design and Implementation (PLDI), June 2011.

Scalable Speculative Parallelization on Commodity Clusters [abstract] (IEEE Xplore, PDF)
Hanjun Kim, Arun Raman, Feng Liu, Jae W. Lee, and David I. August
Proceedings of the 43rd IEEE/ACM International Symposium on Microarchitecture (MICRO), December 2010.
Highest ranked paper in double-blind review process.

Speculative Parallelization Using Software Multi-threaded Transactions [abstract] (ACM DL, PDF)
Arun Raman, Hanjun Kim, Thomas R. Mason, Thomas B. Jablin, and David I. August
Proceedings of the Fifteenth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), March 2010.

Refereed Workshop Publications

Liberty Queues for EPIC Architectures [abstract] (PDF)
Thomas B. Jablin, Yun Zhang, James A. Jablin, Jialu Huang, Hanjun Kim, and David I. August
Proceedings of the Eighth Workshop on Explicitly Parallel Instruction Computer Architectures and Compiler Technology (EPIC), April 2010.

Refereed International Conference Poster

JAWS: A JavaScript Framework for Adaptive CPU-GPU Work Sharing [abstract] (ACM DL, PDF)
Xianglan Piao, Channoh Kim, Younghwan Oh, Huiying Li, Jincheon Kim, Hanjun Kim, and Jae W Lee
Proceedings of the 20th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming - Poster (PPoPP Poster), February 2015.

Efficient CPU-GPU Work Sharing for Data-parallel JavaScript Workloads [abstract] (ACM DL, PDF)
Xianglan Piao, Channoh Kim, Younghwan Oh, Hanjun Kim, and Jae W Lee
Proceedings of the Companion Publication of the 23rd International Conference on World Wide Web Companion (WWW Companion), April 2014.

Refereed Domestic Publications

GPU workload balancing for CSR graph representation [abstract]
Sungjun Cho and Hanjun Kim
Proceedings of the Korea Software Congress, December 2019.

Data Plane Optimization with Pipeline Parallelism [abstract]
Seungbin Song and Hanjun Kim
Proceedings of the Korea Software Congress, December 2018.

Neural Network Optimization for GPU-based Deep Learning Applications Using Weight Matrix Rearrangement [abstract]
Juhyun Kim and Hanjun Kim
Proceedings of the KISS conference, December 2016.

A Survey on Automatic Parallelism [abstract]
Kyoungju Sim and Hanjun Kim
Communications of the Korean Institute of Information Scientists and Engineers, Volume 32, May 2014.

Book Chapters

Automatic Extraction of Parallelism from Sequential Code
David I. August, Jialu Huang, Thomas B. Jablin, Hanjun Kim, Thomas R. Mason, Prakash Prabhu, Arun Raman, and Yun Zhang
Fundamentals of Multicore Software Development (ISBN: 978-1439812730)
Edited by Ali-Reza Adl-Tabatabai, Victor Pankratius, and Walter Tichy. Chapman & Hall / CRC Press, December 2011.

Theses

Context-Aware Memory Dependence Profiling [abstract] (PDF)
Juhyun Kim
Master's Thesis, Department of Computer Science and Engineering, Pohang University of Science and Technology, February 2017.

jSTM: JavaScript Software Transactional Memory System [abstract] (PDF)
Kyoungju Sim
Master's Thesis, Department of Creative IT Engineering, Pohang University of Science and Technology, February 2015.

ASAP: Automatic Speculative Acyclic Parallelization for Clusters [abstract] (PDF)
Hanjun Kim
Ph.D. Thesis, Department of Computer Science, Princeton University, September 2013.