Ilia Petrov

Prof. Ilia Petrov is a faculty member at Reutlingen University's Faculty of Computer Science and Engineering (INF), leading the Data-Management and -Analysis department. He holds the academic rank of Professor and has been affiliated with the university since 2012. His research focuses on advancing data management and analytics, particularly in the realms of database systems, hardware-optimized storage, and near-data processing. He leads the Daten-Management Labor (Data Management Lab), which explores innovations in computational storage, flash memory optimization, and high-performance database architectures.

His research interests include the development of novel storage technologies, such as native computational storage and processing-in-memory systems, alongside efficient indexing methods (e.g., B-Tree variants) for multi-version databases. He has contributed to projects like DANSEN and nativeNDP, which aim to leverage modern hardware for accelerating database operations. Key themes in his work involve minimizing write amplification on flash storage, optimizing HTAP workloads, and enabling efficient near-data processing through FPGA and NVMe integration.

Prof. Petrov's publications span topics including computational storage architectures, Bloom filter optimizations for range queries, and concurrency control in large-scale multi-socket systems. His work emphasizes the synergy between database software and hardware innovations, with applications in enterprise systems and big data analytics. He has collaborated on frameworks like NVMulator for non-volatile memory emulation and Cache-coherent shared locking mechanisms for smart storage systems.

His lab's activities include the development of KV-store accelerators (e.g., nKV) and cost models for query optimization in key-value systems. He has also explored blockchain data structures and snapshot isolation techniques for database transactions. Current trends in his research address the challenges of real-time data processing, cross-layer data formats, and the integration of emerging storage technologies like computational storage drives.