Users require a way to create paretos to compare CAGRA acceleration in lucene with HNSW.
Integrating https://github.com/rapidsai/cuvs-lucene (CPU: lucene-HNSW and GPU: Lucene-CAGRA-HNSW) with cuvs-bench would help create a single source of truth for benchmarks across cuvs integrations.
We will also use this cuvs bench run to write the cuvs-lucene blog and share results with users.