Mixtera: A Data Plane for Foundation Model Training
Summary: Mixtera: declarative data plane for foundation-model training, expressing sample mixtures and visitation order over arbitrary data properties atop existing collections. Centralized read-only layer supports dynamic, feedback-driven reweighting; scales to 256 GH200, no training bottleneck, and implements ADO. (summarized by gpt-5-mini on Apr 11 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Maximilian Böther (ETH Zurich)
- 2. Xiaozhe Yao (ETH Zurich)
- 3. Tolga Kerimoglu (ETH Zurich)
- 4. Dan Graur (ETH Zurich)
- 5. Viktor Gsteiger (ETH Zurich)
- 6. Ana Klimovic (ETH Zurich)
BibTeX Citation
@inproceedings{bother_sigmod26,
title = {{Mixtera: A Data Plane for Foundation Model Training}},
author = {Böther, Maximilian and Yao, Xiaozhe and Kerimoglu, Tolga and Graur, Dan and Gsteiger, Viktor and Klimovic, Ana},
series = {{SIGMOD} '26},
booktitle = {Proceedings of the {ACM} {SIGMOD} International Conference on Management of Data},
publisher = {Association for Computing Machinery},
doi = {10.1145/3786668},
url = {https://dl.acm.org/doi/10.1145/3786668},
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 9 of 9 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 71 | DuckDB: an Embeddable Analytical Database | 2019 | SIGMOD | 0.00037720227 |
| 325 | The Dataflow Model: A Practical Approach to Balancing Correctness, Latency, and Cost in Massive-Scale, Unbounded, Out-of-Order Data Processing | 2015 | VLDB | 0.00020964941 |
| 1,475 | Analyzing and Mitigating Data Stalls in DNN Training | 2021 | VLDB | 0.00010556672 |
| 2,053 | tf.data: A Machine Learning Data Processing Framework | 2021 | VLDB | 9.1140407e-05 |
| 2,246 | PyTorch FSDP: Experiences on Scaling Fully Sharded Data Parallel | 2023 | VLDB | 8.7637336e-05 |
| 4,780 | Data-Juicer: A One-Stop Data Processing System for Large Language Models | 2024 | SIGMOD | 6.4162513e-05 |
| 9,067 | TensorSocket: Shared Data Loading for Deep Learning Training | 2026 | SIGMOD | 5.2283159e-05 |
| 9,069 | Unveiling Challenges for LLMs in Enterprise Data Engineering | 2026 | VLDB | 5.2283159e-05 |
| 9,070 | Scheduling Data Processing Pipelines for Incremental Training on MLP-based Recommendation Models | 2025 | SIGMOD | 5.2283159e-05 |
Previous
Page 1 / 1
Next
Semantically Similar Papers
| # | Overall Rank | Paper | Year | Venue |
|---|---|---|---|---|
| 1 | 1,152 | Cerebro: A Data System for Optimized Deep Learning Model Selection | 2020 | VLDB |
| 2 | 2,053 | tf.data: A Machine Learning Data Processing Framework | 2021 | VLDB |
| 3 | 8,812 | FlexMoE: Scaling Large-scale Sparse Pre-trained Model Training via Dynamic Device Placement | 2023 | SIGMOD |
| 4 | 7,385 | Towards Foundation Database Models | 2025 | CIDR |
| 5 | 11,282 | LobRA: Multi-tenant Fine-tuning over Heterogeneous Data | 2025 | VLDB |
| 6 | 9,144 | Cerebro: A Layered Data Platform for Scalable Deep Learning | 2021 | CIDR |
| 7 | 1,691 | MISTIQUE: A System to Store and Query Model Intermediates for Model Diagnosis | 2018 | SIGMOD |
| 8 | 8,887 | mLoRA: Fine-Tuning LoRA Adapters via Highly-Efficient Pipeline Parallelism in Multiple GPUs | 2025 | VLDB |
| 9 | 10,435 | DFLOP: A Data-driven Framework for Multimodal LLM Training Pipeline Optimization | 2026 | SIGMOD |
| 10 | 4,780 | Data-Juicer: A One-Stop Data Processing System for Large Language Models | 2024 | SIGMOD |