Martin Rupp

Working on a Database system in Java based on Hadoop (Apache HBase + Apache Spark), mainly in external Tables (Parquet) performance optimizations, implementing MERGE INTO SQL Feature, benchmarks, improving code quality and test coverage. Improved, refactored the ODBC Driver (C++, CMake).

Software Engineer for SAP HANA Vora. Mainly C++ Coding, Test Driven Development. Project Lead for a small Team + some students working on "Importer" Topics: Developing efficient readers for BigData File Formats (Parquet, ORC, CSV), BigData File Systems (HDFS, MapR-FS, S3, ...).

Improving tools for simulation (of partial differential equations) in research & development. This included coding and running parallel programs in C++/MPI which ran on germanys biggest clusters. My task is designing the linear algebra and parallel algebraic multigrid. For this I did a lot of speed optimization, also in CUDA and OpenCL. For data analysis and debugging I also wrote a couple of programs in Java, as well as Eclipse and Netbeans plugins for code assistance of our LUA extension

Studying parallel adaptive algebraic multigrid methods.

Created software for documentation and reduction of fertilizers on mobile devices with GPS (Palm OS, Windows CE/Windows Mobile, iOS), software for synchronization of mobile databases with webservers, server software for the calculation of infestation maps.

partielle Differentialgleichungen, wissenschaftliches Rechnen, diskrete/lineare Optimierung, Simulationen auf Großrechnern