P-MOSS: Scheduling Main-Memory Indexes Over NUMA Servers Using Next Token Prediction

Summary: P-MOSS learns NUMA-aware query/core/data placement for main-memory B+-trees, using PMU statistics plus a Decision Transformer framed as next-token prediction. Unlike hand-tuned schedulers, it adapts across hardware/workloads and reports up to 6x throughput over traditional schedules. (summarized by gpt-5-mini on Apr 11 2026)

Paper ID: h8f7447eb492e3fc9
Venue: SIGMOD
Year: 2026
Pagerank: 4.9605343e-05
Overall Rank: 10,738 | 28.03%
DOI: 10.1145/3786675
PDF: Download (CC BY 4.0)

Incoming Non-self Citations Over Time

No non-self incoming citations found for this paper in this database.

Authors

1. Yeasir Rayhan (Purdue University)
2. Walid G. Aref (Purdue University)

BibTeX Citation

@inproceedings{rayhan_sigmod26,
        title = {{P-MOSS: Scheduling Main-Memory Indexes Over NUMA Servers Using Next Token Prediction}},
        author = {Rayhan, Yeasir and Aref, Walid G.},
        series = {{SIGMOD} '26},
        booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
        publisher = {Association for Computing Machinery},
        doi = {10.1145/3786675},
        url = {https://dl.acm.org/doi/10.1145/3786675},
        year = {2026}
}

Incoming Citations (Sorted by Pagerank)

Showing 0 of 0 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank

Outgoing Citations (Sorted by Pagerank)

Showing 21 of 21 cited papers.

Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.

Rank	Cited Paper	Year	Venue	Pagerank
111	DBMSs On A Modern Processor: Where Does Time Go?	1999	VLDB	0.00032621499
215	Morsel-Driven Parallelism: A NUMA-Aware Query Evaluation Framework for the Many-Core Age	2014	SIGMOD	0.00024554257
250	Multi-Core, Main-Memory Joins: Sort vs. Hash Revisited	2014	VLDB	0.00023080446
423	Massively Parallel Sort-Merge Joins in Main Memory Multi-Core Database Systems	2012	VLDB	0.00018450254
1,205	NUMA-aware algorithms: the case of data shuffling	2013	CIDR	0.00011512846
1,269	An Experimental Comparison of Thirteen Relational Equi-Joins in Main Memory	2016	SIGMOD	0.00011236561
1,328	Building a Bw-Tree Takes More Than Just Buzz Words	2018	SIGMOD	0.00011013165
1,391	How Soccer Players Would do Stream Joins	2011	SIGMOD	0.00010790581
1,976	OLTP on Hardware Islands	2012	VLDB	9.2695591e-05
2,147	Quickstep: A Data Platform Based on the Scaling-Up Approach	2018	VLDB	8.9595129e-05
2,400	Parallel R-trees	1992	SIGMOD	8.5059093e-05
3,860	Deployment of Query Plans on Multicores	2015	VLDB	6.9512283e-05
4,036	Scaling Up Concurrent Main-Memory Column-Store Scans: Towards Adaptive NUMA-aware Data and Task Placement	2015	VLDB	6.8259759e-05
4,096	Self-Tuning Query Scheduling for Analytical Workloads	2021	SIGMOD	6.7949494e-05
4,707	Adaptive NUMA-aware data placement and task scheduling for analytical workloads in main-memory column-stores	2017	VLDB	6.4422808e-05
4,797	LSched: A Workload-Aware Learned Query Scheduler for Analytical Database Systems	2022	SIGMOD	6.3978116e-05
5,260	Multi-Disk B-trees	1991	SIGMOD	6.1939544e-05
6,020	Clustering Techniques for Minimizing External Path Length	1996	VLDB	5.8934229e-05
6,332	Robust Performance of Main Memory Data Structures by Configuration	2020	SIGMOD	5.7934902e-05
7,180	Adaptive Hybrid Indexes	2022	SIGMOD	5.5767825e-05
10,104	BUZZARD: A NUMA-Aware In-Memory Indexing System	2013	SIGMOD	5.068372e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	9,191	Database Perspective on LLM Inference Systems	2025	VLDB
2	423	Massively Parallel Sort-Merge Joins in Main Memory Multi-Core Database Systems	2012	VLDB
3	4,461	Memory Management Techniques for Large-Scale Persistent-Main-Memory Systems	2017	VLDB
4	5,603	Database Processing-in-Memory: An Experimental Study	2020	VLDB
5	11,505	LASER: Buffer-Aware Learned Query Scheduling in Master-Standby Databases	2025	VLDB
6	9,011	Improving DBMS Scheduling Decisions with Accurate Performance Prediction on Concurrent Queries	2025	VLDB
7	7,436	Managing Memory For Real-Time Queries	1994	SIGMOD
8	12,863	Adaptively Parallelizing Distributed Range Queries	2009	VLDB
9	4,707	Adaptive NUMA-aware data placement and task scheduling for analytical workloads in main-memory column-stores	2017	VLDB
10	215	Morsel-Driven Parallelism: A NUMA-Aware Query Evaluation Framework for the Many-Core Age	2014	SIGMOD