AI agents & workflows · 2026

Multi-Agent Orchestration for High-Throughput Materials Screening on a Leadership-Class System

Thang Duc Pham, Harikrishna Tummalapalli, Fakhrul Hasan Bhuiyan, Álvaro Vázquez-Mayagoitia, Christine Simpson, Riccardo Balin, Venkatram Vishwanath, Murat Keçeli

arXiv:2604.07681 (2026)

preprint arXiv PDF

Summary

A hierarchical planner–executor multi-agent framework removes the serialization bottleneck of single-agent LLM workflows and orchestrates high-throughput materials-screening campaigns across an exascale (Aurora) system.

Keywords

multi-agent systemsLLM agentshigh-throughput screeningexascaleAuroraworkflow orchestrationAI for HPC

Cite

@article{2026_pham_multiagent_materials_screening,
  title = {Multi-Agent Orchestration for High-Throughput Materials Screening on a Leadership-Class System},
  author = {Thang Duc Pham and Harikrishna Tummalapalli and Fakhrul Hasan Bhuiyan and Álvaro Vázquez-Mayagoitia and Christine Simpson and Riccardo Balin and Venkatram Vishwanath and Murat Keçeli},
  journal = {arXiv:2604.07681 (2026)},
  year = {2026},
  eprint = {2604.07681},
  archivePrefix = {arXiv}
}