Unlock the Potential of Vector Search with Neo4j Aura

We’re excited to introduce a new memory-optimized configuration for Neo4j Aura, purpose-built to enhance vector search. This latest update is designed to help you fast, efficient query results, with even larger datasets.
What’s new?
Neo4j Aura now offers an optional configuration that reallocates memory to prioritize your vector index, enabling:
More data, same performance:
 Load up to twice as much data without compromising query speed.
Smaller instance, same capability:
 Achieve consistent performance with a smaller instance size.
Whether you're looking to scale your dataset or improve efficiency, this configuration ensures your vector search needs are met with reliable performance and flexibility.
Why it matters
Performance in vector search relies on memory availability for indexing. By reserving additional memory for the vector index, Neo4j Aura helps you navigate the performance challenges associated with large datasets. Early testing has shown steady performance up to the point there’s simply no more storage available on disk for more vectors.
Getting started
The vector-optimized configuration is now available for instances with 4GB or more of memory. Switching to this option is as simple as editing the configuration of your existing instance, or setting it when creating a new instance
P.S. Curious about the technical details? Dive deeper into how vector-optimized configurations work and the benefits they unlock here