back to list

Research Project: AvantGraph


AvantGraph is a graph query processing engine under development by the Database group at TU Eindhoven. Designed for efficient processing of both subgraph matching and navigational graph queries, AvantGraph encompasses innovation in three key areas: the planner, the cardinality estimator, and the execution engine.

AvantGraph is designed to be a high-performance, minimum- dependency processing engine for analytical queries over property graphs. The engine is implemented in modern C++ and employs low-level optimizations that reduce performance degradation due to lack of locality, branch mispredictions and non-uniform memory access. AvantGraph is a polyglot engine supporting inputs in both PGM and RDF5 data models. For queries, feature-subsets of Cypher and SPARQL query languages are supported.

Principal Investigator
Nick Yakovets
Involved members:
George Fletcher
Thomas Mulder
Bram van de Wall
Wilco van Leeuwen
Learn more