[DOC] On Disk x4 Compression using Faiss

[naveentatikonda]

What do you want to do?

• Request a change to existing documentation
• Add new documentation
• Report a technical problem with the documentation
• Other

Tell us about your request. Provide a summary of the request.
We have introduced disk based vector search in 2.17 which is providing good customer experience by significantly reducing the costs by dynamically compressing(quantizing) the vectors based on their distribution, providing very good recall with a very minimal increase in query latency. On disk uses Faiss engine for all the compression levels except for x4 (byte quantization), which uses Lucene engine. To obtain better performance for large scale workloads, to keep it consistent with other compression levels and to provide good recall (> 0.95) we will be adding x4 compression support using Faiss engine and setting it as default engine for x4.

Version: List the OpenSearch version to which this issue applies, e.g. 2.14, 2.12--2.14, or all.
2.20

What other resources are available? Provide links to related issues, POCs, steps for testing, etc.
opensearch-project/k-NN#1723