- IBM Research Technical Reports
- Publications via DBLP
- Citations via Google Scholar
- Workshop on Accelerating Data Management Systems Using Modern Processor and Storage Architectures
I am a member of the Systems Acceleration department at the IBM T. J. Watson Research Center. Prior to joining IBM Research in September 1998, I was a post-doctoral fellow at the Center for Advanced Computing Research, California Institute of Technology. I received my PhD in Computer Engineering from Syracuse University. During the graduate studies, my focus was on parallel compilers and parallel out-of-core problems. My PhD work explored compiler and runtime approaches for optimizing parallel out-of-core problems.
I study interactions between applications, programming languages/runtime systems, and computer architectures. I am interested in understanding how modern hardware, multi-core processors, GPUs, and SSDs impact design of optimal algorithms for main-memory and out-of-core problems. At IBM Research, my past projects include Multi-heap Multi-process JVMs, Java Garbage Collection, XML Processing and Parallelization, Cell and GPGPU applications.
My current interest is exploring software-hardware co-design of analytics workloads. I work at the intersection of high-performance computing, analytics, and data management domains. Specifically, I have been investigating how GPUs could be used for accelerating key analytics kernels in text analytics, data management, graph analytics, and deep learning. I collaborate closely with the IBM Toronto Lab, IBM Power Systems, and various Analytics and Database product teams (e.g., Watson, Cognos TM1, Algorithmics, DB2/Netezza, ILOG, and SPSS).
My research agenda can broadly be classified into two main efforts: (1) Acceleration of Large-scale Analytics Systems using holistic software-hardware approaches, and (2) Development of Scalable Concurrent Data-structures on GPU-based hybrid systems (e.g., my work on Lock-free Concurrent Hashtable on GPUs at the Nvidia GTC'14 (Presentation Link)).
I have presented a tutorial on "Analyzing Analytics" at ISCA, PPoPP, and ASPLOS conferences. Tutorial outline can be accessed here. Please contact me if you would like get the slides.
Extended version of this tutorial was published by Morgan Claypool publishes in the Synthesis Lectures of Computer Architecture Series: Analyzing Analytics (November 2015).
For the past few years, I have been organizing a workshop on Accelerating Data Management Systems (ADMS) at the annual VLDB conference (www.adms-conf.org).
I am serving as the Program Vice-Chair for the Distributed Data Management track for the International Conference on Distributed Computing Systems, ICDCS'15 (http://icdcs-2015.cse.ohio-state.edu). I am also on the program committee of the PPoPP'15 conference.