Yogish Sabharwal

Contact Information:

email: yogish@cse.iitd.ernet.in

Workplace:
IBM India Research Lab
phone: +91-11-41292195
email: ysabharwal@in.ibm.com


I completed my Ph.D. from the Indian Institute of Technology, Delhi in 2007. My supervisors were Prof. Sandeep Sen and Prof. Amit Kumar.
My thesis is in approximation algorithms for proximity and clustering problems.

I work in the High Performance Computing (HPC) group at IRL. Amongst other things, I have worked on optimization of HPC Challenge (HPCC) benchmarks on IBM Supercomputers such as Blue Gene. These benchmarks are used in evaluating the performance of HPC systems. Our optimizations of the RandomAccess benchmark have won the HPCC Class 1 award for this benchmark on IBM's Blue Gene Supercomputers five years in a row (2005-2009). Associated publications include a best paper finalist at Supercomputing 2006 and best paper award at IPDPS 2009.

Research Interests

    Computational Geometry, Approximation Algorithms, High Performance Computing

Publications

  • Computational Geometry, Approximation Algorithms

    • Linear time approximate clustering in any dimension
      with Amit Kumar and Sandeep Sen
      Journal of the ACM (JACM) - to appear.

    • Approximating Decision Trees with Multiway Branches
      with Venkatesan T. Chakaravarthy, Vinayaka Pandit and Sambuddha Roy
      International Colloquium on Automata, Languages and Programming (ICALP) 2009.

    • Analysis of sampling techniques for association rule mining
      with Vinayaka Pandit and Venkatesan T. Chakaravarthy,
      International Conference on Database Theory (ICDT) 2009.

    • The Priority k-median problem
      with Amit Kumar,
      Foundations of Software Technology and Theoretical Computer Science (FSTTCS) 2007.

    • Nearest Neighbor Searching using Point Location in Balls with applications to Approximate Voronoi Diagrams
      with Nishant Sharma and Sandeep Sen,
      Journal of Computer and System Sciences (JCSS). Vol 72(6), 2006.

    • A Simple Linear time randomized algorithm for approximate 2-means clustering
      with Sandeep Sen,
      Computational Geometry: Theory and Applications (CGTA), Vol 32(2), 2005.

    • Linear Time Algorithms for Clustering Problems in any Dimensions
      with Amit Kumar and Sandeep Sen,
      International Colloquium on Automata, Languages and Programming (ICALP) 2005.

    • A Simple linear time (1+ε)-approximation algorithm for k-means clustering in any dimensions
      with Amit Kumar and Sandeep Sen,
      Foundations of Computer Science (FOCS) 2004.

    • Nearest Neighbor Searching using Point Location in Balls with applications to Approximate Voronoi Diagrams
      with Nishant Sharma and Sandeep Sen,
      Foundations of Software Technology and Theoretical Computer Science (FSTTCS) 2002.

  • High Performance Computing

    • Efficient Algorithms for Global Snapshots in Large Distributed Systems
      with Rahul Garg and Vijay K. Garg
      IEEE Transactions on Parallel and Distributed Systems (to appear).

    • Varying Bandwidth Resource Allocation Problem with Bag Constraints
      with Venkatesan T. Chakaravarthy, Vinayaka Pandit and Deva P. Seetharam
      International Parallel and Distributed Processing Symposium (IPDPS) 2010.

    • HPCC RandomAccess Benchmark for Next Generation Supercomputers
      with Vikas Aggarwal, Rahul Garg and Philip Heidelberger
      International Parallel and Distributed Processing Symposium (IPDPS) 2009 {Best Paper award winner}.

    • Optimization of Fast Fourier Transforms on the Blue Gene/L Supercomputer
      with Saurabh K. Garg, Rahul Garg, John A. Gunnels and Ramendra K. Sahoo
      International Conference on High Performance Computing (HiPC) 2008.

    • Optimization of BLAS on the Cell Processor
      with Vaibhav Saxena, Prashant Agrawal, Vijay K. Garg, Vimitha Kuruvilla and John A. Gunnels
      International Conference on High Performance Computing (HiPC) 2008.

    • Performance Analysis and Optimization of All-to-all communication on the Blue Gene/L Supercomputer
      with Sameer Kumar, Rahul Garg and Philip Heidelberger
      International Conference on Parallel Processing (ICPP) 2008.

    • Optimizations in Financial Engineering : The Least-Squares Monte Carlo method of Longstaff and Schwartz
      with Anamitra Choudhury, Alan King and Sunil Kumar
      International Parallel and Distributed Processing Symposium (IPDPS) 2008.

    • Scalable Algorithms for Global Snapshots in Distributed Systems
      with Rahul Garg and Vijay K. Garg,
      International Conference on Supercomputing (ICS) 2006.

    • Optimizing the HPCC Randomaccess Benchmark on Blue Gene/L Supercomputer
      with Rahul Garg,
      International Conference on Measurement and Modeling of Computer Systems (SIGMETRICS) 2006 {Poster Paper}.

    • Software Routing and Aggregation of Messages to Optimize the Performance of the HPCC Randomaccess Benchmark
      with Rahul Garg,
      Supercomputing (SC) 2006 {Best Paper award finalist}.

    • Large Scale Drop Impact Analysis of Mobile Phone Using ADVC on Blue Gene/L
      with H. Akiba, T. Ohyama, Y. Shibata, K. Yuyama, Y. Katai, R. Takeuchi, T. Hoshino, S. Yoshimura, H. Noguchi, M. Gupta, J. Gunnels, V. Austel, R. Garg, S. Kato, T. Kawakami, S. Todokoro and J. Ikeda
      Supercomputing (SC) 2006 {Gordon Bell award finalist}.

Invited Talks

  • Randomized Techniques in Geometry [pdf],
    Workshop on Introduction to Geometric Algorithms, IIT Kharagpur, 2008.

  • Writing efficient parallel programs for next-generation supercomputers,
    User Industry Symposium on Parallel Application Development through MPI, International Conference on High Performance Computing (HiPC) 2008.