systems research at HP Labs

Taught CS I and II, computer systems, and upper-level electives in systems.

Large-scale data analytics on billions of daily data items. Responsibilities included scripting of data collection; number crunching using Hadoop / Hive; statistical and predictive analysis with python and R; and graphical and verbal communication of analysis.

Designed, benchmarked, and optimized software for various hardware technologies, including conventional and non-conventional CPUs, GPGPUs and NICs.

Developed software for accelerating Web search (Powerset/Bing technologies).

R&D in scalability of large-scale web servers. Developed with accelerator architectures and advanced algorithms for high-performance semantic search.

Researched advanced parallel job scheduling techniques and novel system software architectures. Developed and evaluated a successful job scheduling algorithm (FCS) and participated in developing a novel model for scalable system software, using novel interconnection technology (Quadrics). Additional research topics included network protocols and resource management techniques. Work resulted in over 20 peer-reviewed publications.

Researched job scheduling and network allocation strategies for BlueGene/L, using analytical and simulated evaluations. Work resulted in patent application on allocation algorithms.

Developed communication algorithms for data transport over RF links for Tadiran Communications.Developed data compression algorithm for IP-over-satellite for Harmonic Inc. (USPTO application #20030030575)

Led software development teams on three scientific computing projects. Work included various aspects of high-performance computation, optimization, and physics simulations, and resulted in successful and widely-used applications.