Elastic N.V. announced new performance and cost-efficiency breakthroughs for its vector search capabilities with the introduction of ACORN, a smart filtering algorithm, and Better Binary Quantization (BBQ) as the default for high-dimensional dense vectors. These enhancements improve query performance and ranking quality.
ACORN-1, a new algorithm for filtered k-Nearest Neighbor (kNN) search in Elasticsearch, delivers up to 5X speedups in real-world filtered vector search benchmarks. It tightly integrates filtering into the HNSW graph traversal, allowing flexible filter definition at query time without compromising result accuracy.
Better Binary Quantization (BBQ) is now the default quantization method for dense vectors of 384+ dimensions in Elasticsearch 9.1. This change boosts ranking quality while significantly reducing latency and resource usage, achieving approximately 32X compression. BBQ outperformed traditional float32-based search in 9 out of 10 industry-standard BEIR datasets, maintaining high ranking accuracy.
The content on BeyondSPX is for informational purposes only and should not be construed as financial or investment advice. We are not financial advisors. Consult with a qualified professional before making any investment decisions. Any actions you take based on information from this site are solely at your own risk.