Built with and by Teycir Ben Soltane•
How to Use•FAQ•GitHub•arXiv.org•
Share:
ArXivCSExplorer
☆☆Bookmarks🏆RSSHow to UseFAQ
Home/Authors/Teng Su

Teng Su

2 indexed papers

Recent (6 mo)
2
With code
0
Influential cites
0
Benchmarked
0

Publications per year

2
26

Top categories

Distributed×2

Frequent co-authors

Mo Sun1×
Yifan Yao1×
Yanwei Liu1×
Luobin Liu1×
Zhenzhang Yang1×
Kaisheng Wang1×

Research Timeline

2026
HyperParallel-Mpipe: A Composable Algebra System for Optimizing MLLM Training over Supernode Clusters

The paper introduces Mpipe, a method for multimodal-aware heterogeneous parallel scheduling in large-scale multimodal language model training, achieving significant speedups on Ascend 910C NPU clusters.

HyperParallel-FSDP: Topology-Aware Fully Sharded Training with Layout-Driven Muon on Ascend SuperPods

The paper presents HyperParallel-FSDP, a parallelization framework that improves scalability and efficiency for large-scale deep learning models.

Highlighted terms show continued research focus across papers

Papers

cs.DCEmpiricalRecentSep 18, 2026

HyperParallel-FSDP: Topology-Aware Fully Sharded Training with Layout-Driven Muon on Ascend SuperPods

Mo Sun, Yifan Yao, Yanwei Liu, Luobin Liu +13 more

The paper presents HyperParallel-FSDP, a parallelization framework that improves scalability and efficiency for large-scale deep learning models.

View →
cs.DCEmpiricalRecent
Jul 3, 2026

HyperParallel-Mpipe: A Composable Algebra System for Optimizing MLLM Training over Supernode Clusters

Chong Li, Zhengdao Yu, Nelson Lossing, Thibaut Tachon +5 more

The paper introduces Mpipe, a method for multimodal-aware heterogeneous parallel scheduling in large-scale multimodal language model training, achieving significant speedups on Ascend 910C NPU cluster…

View →