# Vector store optimization
<a name="genperf04"></a>


| GENPERF04: How do you improve the performance of data retrieval systems? | 
| --- | 
|   | 

 Data retrieval systems like vector databases support some of the most popular design patterns for generative AI systems. A performance bottleneck in a data retrieval system can have cascading downstream effects, which are difficult to identify. A thorough understanding of data embedding and retrieval systems can help mitigate downstream performance issues. Ultimately, a thorough understanding of the kind of data being tokenized and queried, as well as data access patterns, can help reduce performance issues in the long-run. 

**Topics**
+ [GENPERF04-BP01 Test vector embeddings for latency and relevant performance](genperf04-bp01.md)
+ [GENPERF04-BP02 Optimize vector sizes for your use case](genperf04-bp02.md)