Taesoo Kim’s Systems Group Has Strong Showing at SOSP

SOSP location

School of Computer Science researchers have five papers accepted at one of the main systems conferences. Associate Professor Taesoo Kim’s  Systems Software & Security Lab group contributed four of them, showing their strength in the field.

The ACM Symposium on Operating Systems Principles (SOSP) is the premier conference for operating systems researchers. Now in its 27th year, the conference was held in Huntsville, Ontario, Canada, this year from October 27 to 30.

“With five accepted papers in a single year, this puts Georgia Tech as one of the best, active research group in systems,” Kim said.

Georgia Tech papers contributed to work on file systems, including:


  • Scalable and Practical Locking with Shuffling
    Sanidhya Kashyap(Georgia Tech), Irina Calciu (VMware Research Group), Xiaohe Cheng (Hong Kong University of Science and Technology), Changwoo Min (Virginia Tech), Taesoo Kim (Georgia Tech)

    High-performance multicore system software is built on lock algorithms. This research identifies four performance factors in locks and offers a new technique to manage all these factors without slowing down the lock.

  • Recipe: Converting Concurrent DRAM Indexes to Persistent-Memory Indexes
    Se Kwon Lee and Jayashree Mohan (University of Texas at Austin), Sanidhya Kashyap and, Taesoo Kim (Georgia Tech), Vijay Chidambaram (University of Texas at Austin and VMware Research)

    Recipe is an approach to convert concurrent DRAM indexes into indexes that work in crashes for persistent memory.


  • SplitFS: Reducing Software Overhead in File Systems for Persistent Memory
    Rohan Kadekodi and Se Kwon Lee (University of Texas at Austin), Sanidhya Kashyap and  Taesoo Kim (Georgia Tech), Aasheesh Kolli (Penn State University and VMware Research), Vijay Chidambaram (University of Texas at Austin and VMware Research)

    This file system for persistent memory reduces software overhead by presenting a novel split between user-space library file systems and existing kernel persistent memory file systems.


  • Lineage Stash: Fault Tolerance Off the Critical Path
    Stephanie Wang (UC Berkeley), John Liagouris (ETH Zurich), Robert Nishihara (UC Berkeley), Philipp Moritz (UC Berkeley), Ujval Misra (UC Berkeley), Alexey Tumanov (Georgia Tech), Ion Stoica (UC Berkeley)

    Lineage stash is a decentralized casual logging technique that helps cluster computing tolerate failures.

Core Research Areas: 

Tess Malone, Communications Officer