The Vector Search Edition includes a specification calculator to estimate resource requirements for an instance. You can use it to plan your budget before creating a new instance.
Use the calculator
Click Specification Calculator.
The OpenSearch Vector Search Edition specification calculator dialog box opens, containing three configuration sections: Basic configurations, Vector data configurations, and Query configurations. After you configure the parameters, click Perform Operation to get the recommended specifications.
Basic configurations
-
Instance region: The region for the new instance.
-
Disaster recovery: Select Yes or No to specify if disaster recovery is required.
Vector data configurations
-
Number of vector docs: The total number of vector documents to be stored in the instance.
-
Vector dimension: The dimension of the vectors to be stored in the instance.
-
Vector algorithm: Select a vector algorithm based on your requirements. Three algorithms are supported:
-
HNSW: A graph-based vector retrieval algorithm that offers a very high retrieval rate and excellent performance. Its memory and storage consumption is comparable to Linear. HNSW performs well on both low-dimensional and high-dimensional vector datasets, making it suitable for most vector search scenarios.
-
QC: A quantization and clustering-based vector retrieval algorithm. It offers a good balance of retrieval accuracy and performance, consumes fewer resources, and provides good performance. It performs better on low-dimensional vector datasets. Its memory and storage consumption is typically about one-fourth of that required by Linear and HNSW. QC is suitable for large-scale retrieval scenarios that do not have strict requirements for the retrieval rate.
-
Linear: A linear search algorithm, also known as brute-force search. It guarantees completely accurate retrieval results but consumes significant resources and offers lower performance. It is typically used for precise searches on small datasets with up to 10,000 entries.
Query configurations
-
Average QPS: The expected queries per second (QPS).
-
Expected average response time: The target average response time for a query. Select the unit as seconds (s) or milliseconds (ms).
After you complete the configurations, click Perform Operation.
After the calculation is complete, the system recommends the required specifications and number of replicas for the Query Result Searcher (QRS) workers and searcher workers.
The recommendation includes parameters such as QRS worker specification family, number of QRS workers, QRS worker specification, searcher worker specification family, number of searcher workers, searcher worker specification, and disk space per searcher worker.
You can then purchase an instance with the recommended specifications.