Andrew Kane

Hi @aytekinar, please create a new issue for this.

@pashkinelfe That's my understanding as well. It'd be great if someone wanted to explore this. I think the decision to support it would come down to the complexity it adds....

I think we should probably try to update the cost estimate so the index won't be used in cases where few tuples are expected to be returned due to attribute...

Pushed one way to address this in #424 (essentially, the queries in this thread are not a good fit for a vector index, so the PR makes sure one isn't...

Hi @aytekinar, thanks for the PR! It's a neat idea, and the code looks really clean. However, it adds a good amount of complexity. I think a minimal version would...

Hi @ashvardanian, thanks for sharing. Looks really neat, but don't want to take on any external dependencies or different licenses (however, if someone decides to fork and benchmark, I'd be...

@ashvardanian The best way to benchmark is ann-benchmarks. @aytekinar `target_clones` is neat, but looks like it may not be trivial to detect when it's supported. With Apple clang 15 on...

Thanks @aytekinar, looks like CI is now passing. However, I don't think it makes sense to spend more time on this unless benchmarks show significant improvement in overall performance.

A few other places to look for performance gains would be: 1. Index build times 2. IVFFlat index scans 3. Table scans Based on past SIMD optimizations (#180), index build...

Hi all, I'm considering including a version of this in 0.7.0 (and removing `-march=native` by default). I did some initial benchmarking with [this branch](https://github.com/pgvector/pgvector/compare/target-clones) on HNSW build time (single process)...