Oracle outperformed Neo4j in 19 of the 25 queries on the standard Linked Data Benchmark Council Social Network Benchmark. Neo4j had better results in only 4 queries with both products reaching a time-out threshold on 2 queries.
The tests used, the LDBC Social Network Benchmark (SNB)
Faster Query Performance
Oracle Property Graph was up to 250 times faster when compared with published tests on Neo4j performed by University of California-Merced for queries that completed in under five hours. There were three queries that Oracle completed that Neo4j did not.
Oracle Property Graph is not just Fast, but Full of Features
Oracle Property Graph includes a high-performance, in-memory graph server – PGX – to optimize graph query and algorithm execution. This native graph engine enables parallel execution of graph queries and algorithms. It includes nearly 60 prebuilt graph algorithms that can be invoked and applied to both property graphs and RDF graphs; users can also write their own high-performance algorithms using Java and Oracle Property Graph compilers.
In addition, the property graph includes PGQL, a SQL-like native graph query language, a graph visualization component, and notebook support. Both Oracle Property Graph and Oracle RDF Graph are part of Oracle Database, the world’s first converged database. A converged database is a multi-model, multitenant, multi-workload database that supports the data models and access methods required by applications, analysts and developers.
Download the report and read more, or get started with a step-by-step workshop.