IcebergHT: High Performance Hash Tables Through Stability and Low Associativity
Summary: IcebergHT uses stability (no item movement) and low associativity (few candidate slots) to minimize cache-line traffic and enable crash-safe updates. Iceberg hashing with in-memory metadata is space-efficient and fast, delivering PMEM inserts 50%–3× faster and queries 20%–2× faster, with ~17% space overhead. (summarized by gpt-5-nano on Feb 09 2026)
Incoming Non-self Citations Over Time
Authors
- 1. Prashant Pandey (University of Utah)
- 2. Michael A. Bender (State University of New York at Stony Brook)
- 3. Alex Conway (VMware)
- 4. Martín Farach-Colton (Rutgers University)
- 5. William Kuszmaul (Massachusetts Institute of Technology)
- 6. Guido Tagliavini (Rutgers University)
- 7. Rob Johnson (VMware)
BibTeX Citation
@inproceedings{pandey_sigmod23,
title = {{IcebergHT: High Performance Hash Tables Through Stability and Low Associativity}},
author = {Pandey, Prashant and Bender, Michael A. and Conway, Alex and Farach-Colton, Martín and Kuszmaul, William and Tagliavini, Guido and Johnson, Rob},
series = {{SIGMOD} '23},
booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
publisher = {Association for Computing Machinery},
doi = {10.1145/3588727},
url = {https://dl.acm.org/doi/10.1145/3588727},
year = {2023}
}Incoming Citations (Sorted by Pagerank)
Showing 7 of 7 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 9,561 | Global Hash Tables Strike Back! An Analysis of Parallel GROUP BY Aggregation | 2026 | VLDB | 5.1571823e-05 |
| 9,562 | H-Rocks: CPU-GPU accelerated Heterogeneous RocksDB on Persistent Memory | 2025 | SIGMOD | 5.1571823e-05 |
| 9,866 | Zombie Hashing: Reanimating Tombstones in a Graveyard | 2025 | SIGMOD | 5.1176637e-05 |
| 10,569 | Focus! Fast On-disk Concurrency-control Using Sketches | 2026 | SIGMOD | 4.9793485e-05 |
| 10,623 | Breadcrumb Filters: Fast Fully Featured Filters | 2026 | SIGMOD | 4.9793485e-05 |
| 10,777 | Succinct and Fast Tiny Pointer Hash Tables | 2026 | VLDB | 4.9793485e-05 |
| 11,052 | SHARD: A Scalable and Resize-optimized Hash Index on Disaggregated Memory | 2026 | VLDB | 4.9793485e-05 |
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 3 of 3 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 546 | Faster: A Concurrent Key-Value Store with In-Place Updates | 2018 | SIGMOD | 0.00016590738 |
| 1,359 | Dash: Scalable Hashing on Persistent Memory | 2020 | VLDB | 0.00010919021 |
| 4,516 | Persistent Memory Hash Indexes: An Experimental Evaluation | 2021 | VLDB | 6.5646895e-05 |
Previous
Page 1 / 1
Next
Semantically Similar Papers
| # | Overall Rank | Paper | Year | Venue |
|---|---|---|---|---|
| 1 | 4,144 | Memory-Contention Responsive Hash Joins | 1994 | VLDB |
| 2 | 11,052 | SHARD: A Scalable and Resize-optimized Hash Index on Disaggregated Memory | 2026 | VLDB |
| 3 | 9,866 | Zombie Hashing: Reanimating Tombstones in a Graveyard | 2025 | SIGMOD |
| 4 | 2,245 | Cache-Efficient Aggregation: Hashing Is Sorting | 2015 | SIGMOD |
| 5 | 8,241 | Pea Hash: A Performant Extendible Adaptive Hashing Index | 2023 | SIGMOD |
| 6 | 1,283 | A Seven-Dimensional Analysis of Hashing Methods and its Implications on Query Processing | 2016 | VLDB |
| 7 | 14,015 | Cache-Oblivious Hashing | 2010 | PODS |
| 8 | 969 | Memory-Efficient Hash Joins | 2015 | VLDB |
| 9 | 10,777 | Succinct and Fast Tiny Pointer Hash Tables | 2026 | VLDB |
| 10 | 4,516 | Persistent Memory Hash Indexes: An Experimental Evaluation | 2021 | VLDB |