Turbocharging Vector Databases using Modern SSDs
Summary: SSD-aware suite for disk-resident graph-based ANN (HNSW): parallel io_uring I/O, spatially-aware insertion reordering, and locality-preserving colocation, implemented in pgvector. Yields up to 11.1× query throughput, 3.23× cache-hit, 98.4% faster index build. (summarized by gpt-5-mini on Feb 09 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Joobo Shim
- 2. Jaewon Oh
- 3. Hongchan Roh
- 4. Jaeyoung Do
- 5. Sang-Won Lee
Incoming Citations (Sorted by Pagerank)
Showing 0 of 0 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 2 of 2 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 812 | A Modeling Study of the TPC-C Benchmark | 1993 | SIGMOD | 0.00016408423 |
| 2,690 | Starling: An I/O-Efficient Disk-Resident Graph Index Framework for High-Dimensional Vector Similarity Search on Data Segment | 2024 | SIGMOD | 8.293714e-05 |
Previous
Page 1 / 1
Next