Heterogeneity-Aware Distributed Machine Learning Training via Partial Reduce
Summary: Partial-reduce: heterogeneity-aware all-reduce variant for distributed ML; decomposes synchronous all-reduce into parallel-asynchronous blocks to tolerate stragglers. Converges to a stationary point at sublinear SGD rate; adds dynamic, staleness-aware averaging and group-generation to avoid update isolation; prototype yields 1.21x–2x speedups. (summarized by gpt-5-nano on Feb 09 2026)
Incoming Non-self Citations Over Time
Authors
- 1. Xupeng Miao
- 2. Xiaonan Nie
- 3. Yingxia Shao
- 4. Zhi Yang
- 5. Jiawei Jiang
- 6. Lingxiao Ma
- 7. Bin Cui
Incoming Citations (Sorted by Pagerank)
Showing 14 of 14 citing papers.
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 6 of 6 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 411 | PyTorch Distributed: Experiences on Accelerating Data Parallel Training | 2020 | VLDB | 0.00023906921 |
| 1,044 | DimmWitted: A Study of Main-Memory Statistical Analytics | 2014 | VLDB | 0.00014475229 |
| 1,942 | Heterogeneity-aware Distributed Parameter Servers | 2017 | SIGMOD | 0.00010012691 |
| 2,642 | Vertica-ML: Distributed Machine Learning in Vertica Database | 2020 | SIGMOD | 8.3851878e-05 |
| 3,808 | SketchML: Accelerating Distributed Machine Learning with Data Sketches | 2018 | SIGMOD | 6.7455428e-05 |
| 4,964 | PS2: Parameter Server on Spark | 2019 | SIGMOD | 5.7965988e-05 |
Previous
Page 1 / 1
Next