Publications

2026

1. SC’26

   BatchFlow: Shared Batch Reuse and Adaptive Scheduling for Multi-Job Training

   Patrick Watters, Xiaolong Ma, Syed Zawad, and 4 more authors

   In Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC)Acceptance rate: 19% , Nov 2026
2. SC’26 Workshop

   OPSERVE: Opportunistic LLM Inference over Fragmented GPU Capacity in HPC Systems

   Patrick Watters, Xiaolong Ma, Harry Yu, and 5 more authors

   In Proceedings of the AI on HPC Workshop at the International Conference for High Performance Computing, Networking, Storage and Analysis (AIonHPC @ SC), Nov 2026

   Patrick Watters and Xiaolong Ma contributed equally (co-first authors)
3. Preprint

   SOLAR: Communication-Efficient Model Adaptation via Subspace-Oriented Latent Adapter Reparametrization

   Seyed Mahmoud Sajjadi Mohammadabadi, Xiaolong Ma, Lei Yang, and 2 more authors

   2026
4. Preprint

   PEML: Parameter-efficient Multi-Task Learning with Optimized Continuous Prompts

   Anjir Ahmed Chowdhury, Syed Zawad, Xiaolong Ma, and 2 more authors

   2026
5. Preprint

   Wrong but Useful: Trajectory Value Beyond Answer Correctness in Multi-Agent Messages

   Chih-Hsuan Yang, Anjir Ahmed Chowdhury, Cheng-Hau Yang, and 7 more authors

   Aug 2026

2025

1. Preprint

   Diffusion-Based, Data-Assimilation-Enabled Super-Resolution of Hub-height Winds

   Xiaolong Ma, Xu Dong, Ashley Tarrant, and 5 more authors

   Oct 2025
2. PEVA’25

   Enabling Scalable and Adaptive Machine Learning Training via Serverless Computing on Public Cloud

   Xiaolong Ma, Ahsan Ali, Syed Zawad, and 5 more authors

   Performance Evaluation, Mar 2025

   Ahsan Ali and Xiaolong Ma contributed equally (co-first authors)

   DOI
3. PEVA’25

   FedCust: Offloading Hyperparameter Customization for Federated Learning

   Xiaolong Ma, Syed Zawad, Jun Yi, and 5 more authors

   Performance Evaluation, Mar 2025

   Syed Zawad, Xiaolong Ma, and Jun Yi contributed equally (co-first authors)

   DOI

2024

1. ICPE’24

   MalleTrain: Deep Neural Networks Training on Unfillable Supercomputer Nodes

   Xiaolong Ma, Feng Yan, Lei Yang, and 4 more authors

   In Proceedings of the 15th ACM/SPEC International Conference on Performance Engineering (ICPE), May 2024

   Bib

   @inproceedings{ma2024malletrain,
     title = {MalleTrain: Deep Neural Networks Training on Unfillable Supercomputer Nodes},
     author = {Ma, Xiaolong and Yan, Feng and Yang, Lei and Foster, Ian and Papka, Michael and Liu, Zhengchun and Kettimuthu, Rajkumar},
     booktitle = {Proceedings of the 15th ACM/SPEC International Conference on Performance Engineering (ICPE)},
     year = {2024},
     address = {South Kensington, London, UK},
     month = may
   }

2023

1. VLDB’23

   InfiniStore: Elastic Serverless Cloud Storage

   Jingyuan Zhang, Ao Wang, Xiaolong Ma, and 8 more authors

   In Proceedings of the 49th International Conference on Very Large Data Bases (VLDB), Mar 2023

2020

1. FAST’20

   InfiniCache: Exploiting Ephemeral Serverless Functions to Build a Cost-Effective Memory Cache

   Ao Wang, Jingyuan Zhang, Xiaolong Ma, and 6 more authors

   In Proceedings of the 18th USENIX Conference on File and Storage Technologies (FAST)Acceptance rate: 23/138 = 16.7% , Feb 2020