Kirin: Efficient In-Storage Learned Compaction for LSM-Trees via System-Algorithm Co-Design
Summary: Kirin co-designs learned LSM compaction with computational storage: model training is embedded in compaction to hide latency and keep models current, while host–CSD collaboration parallelizes work and reduces indexing I/O. It substantially improves read/write throughput and preserves low read latency. (summarized by gpt-5.6-luna on Aug 17 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Guifeng Wang (Shanghai Jiao Tong University)
- 2. Shengan Zheng (Shanghai Jiao Tong University)
- 3. Penghao Sun (Shanghai Jiao Tong University)
- 4. Jin Pu (Shanghai Jiao Tong University)
- 5. Kaijiang Deng (Shanghai Jiao Tong University)
- 6. Bowen Zhang (Shanghai Jiao Tong University)
- 7. Weihan Kong (Shanghai Jiao Tong University)
- 8. Cong Zhou (Shanghai Jiao Tong University)
- 9. Yifan Hua (Peking University)
- 10. Linpeng Huang (Shanghai Jiao Tong University)
BibTeX Citation
@article{wang_vldb26,
title = {{Kirin: Efficient In-Storage Learned Compaction for LSM-Trees via System-Algorithm Co-Design}},
author = {Wang, Guifeng and Zheng, Shengan and Sun, Penghao and Pu, Jin and Deng, Kaijiang and Zhang, Bowen and Kong, Weihan and Zhou, Cong and Hua, Yifan and Huang, Linpeng},
journal = {PVLDB},
series = {{VLDB} '26},
volume = {19},
number = {9},
pages = {1991--2004},
doi = {10.14778/3819518.3819529},
url = {https://doi.org/10.14778/3819518.3819529},
year = {2026}
}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 14 of 14 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
Previous
Page 1 / 1
Next