



Research paper proposing a decoupled storage architecture for vector databases that improves update speed by 10.05x for insertions and 6.89x for deletions through innovative design.
Loading more......
This research paper published on arXiv (2510.25401) in December 2025 proposes a novel decoupled storage architecture for vector databases that dramatically improves update performance while maintaining query efficiency.
The paper introduces a new architectural approach that separates:
This decoupling enables independent optimization of different workload patterns.
Benchmarked against:
Across datasets:
The paper identifies opportunities for:
Free access on arXiv. Code and datasets planned for release.