PLIN: A Persistent Learned Index for Non-Volatile Memory with High Performance and Instant Recovery
Summary: PLIN adapts learned indexes to NVM by redesigning node layout and placement—locally unordered but globally ordered leaves, model-copying, and hierarchical insertions to minimize NVM block misses and speed updates. NVM-only design with optimistic concurrency yields ~2.08x insert, ~4.42x query speedups over prior NVM indices and ~30 μs instant recovery. (summarized by gpt-5-mini on Feb 09 2026)
Incoming Non-self Citations Over Time
Authors
- 1. Zhou Zhang (University of Science and Technology Beijing)
- 2. Zhaole Chu (University of Science and Technology Beijing)
- 3. Peiquan Jin (University of Science and Technology Beijing)
- 4. Yongping Luo (University of Science and Technology Beijing)
- 5. Xike Xie (University of Science and Technology Beijing)
- 6. Shouhong Wan (University of Science and Technology Beijing)
- 7. Yun Luo (Tencent)
- 8. Xufei Wu (Tencent)
- 9. Peng Zou (Tencent)
- 10. Chunyang Zheng (Intel)
- 11. Guoan Wu (Intel)
- 12. Andy Rudoff (Intel)
BibTeX Citation
@article{zhang_vldb23,
title = {{PLIN: A Persistent Learned Index for Non-Volatile Memory with High Performance and Instant Recovery}},
author = {Zhang, Zhou and Chu, Zhaole and Jin, Peiquan and Luo, Yongping and Xie, Xike and Wan, Shouhong and Luo, Yun and Wu, Xufei and Zou, Peng and Zheng, Chunyang and Wu, Guoan and Rudoff, Andy},
journal = {PVLDB},
series = {{VLDB} '23},
volume = {16},
number = {2},
pages = {243--255},
doi = {10.14778/3565816.3565826},
url = {https://doi.org/10.14778/3565816.3565826},
year = {2023}
}Incoming Citations (Sorted by Pagerank)
Showing 14 of 14 citing papers.
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 19 of 19 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
Previous
Page 1 / 1
Next
Semantically Similar Papers
| # | Overall Rank | Paper | Year | Venue |
|---|---|---|---|---|
| 1 | 2,707 | Evaluating Persistent Memory Range Indexes | 2020 | VLDB |
| 2 | 6,710 | PIM-tree: A Skew-resistant Index for Processing-in-Memory | 2023 | VLDB |
| 3 | 5,993 | NBTree: a Lock-free PM-friendly Persistent B+-Tree for eADR-enabled PM Systems | 2022 | VLDB |
| 4 | 5,423 | Updatable Learned Indexes Meet Disk-Resident DBMS - From Evaluations to Design Choices | 2023 | SIGMOD |
| 5 | 2,246 | DPTree: Differential Indexing for Persistent Memory | 2020 | VLDB |
| 6 | 847 | Benchmarking Learned Indexes | 2021 | VLDB |
| 7 | 948 | BzTree: A High-Performance Latch-free Range Index for Non-Volatile Memory | 2018 | VLDB |
| 8 | 2,910 | APEX: A High-Performance Learned Index on Persistent Memory | 2022 | VLDB |
| 9 | 1,823 | LB+-Trees: Optimizing Persistent Index Performance on 3DXPoint Memory | 2020 | VLDB |
| 10 | 621 | Persistent B+-Trees in Non-Volatile Main Memory | 2015 | VLDB |