Models
Collaborative Foundation models
Federated pretraining and fine-tuning of language, vision, and multimodal models across distributed supercomputing facilities.
In conjunction with IEEE International Conference on Cluster Computing 2026
Bringing together researchers and practitioners shaping collaborative AI across edge, cloud and high performance computing - without compromising data sovereignty.
01 / About
The 3rd Workshop on Federated and Privacy-Preserving AI for HPC (FPAI-HPC'26) brings together researchers and practitioners from academia, industry, government, and research laboratories to advance trustworthy AI across distributed supercomputing environments without centralizing sensitive data. The workshop explores scalable federated training, communication efficiency, privacy and security, foundation models, scientific applications, sustainable AI, and agentic workflows while fostering shared platforms, meaningful evaluation metrics, and new collaborations for AI at HPC scale.
02 / Scope
Building on previous FPAI-HPC workshops at the Supercomputing conference series (SC), this edition brings together academia, industry, federal agencies, and research laboratories to advance federated and privacy-preserving AI for distributed HPC environments. Its scope spans the full stack—from foundation models, scalable aggregation, and communication-efficient systems to security, sustainability, scientific computing, and agentic workflows—connecting foundational theory with deployable systems and real-world applications.
Models
Federated pretraining and fine-tuning of language, vision, and multimodal models across distributed supercomputing facilities.
Scale
Hierarchical federated learning, hybrid parallelism, and workload-aware optimization for multi-cluster systems.
Systems
Methods that bridge fast intra-facility interconnects and slower, higher-latency links between facilities.
Trust
Adversarial robustness, backdoor vulnerabilities, data sovereignty, and differential privacy trade-offs.
Efficiency
Energy-aware resource allocation that maintains accuracy for federated AI at HPC scale.
Science
Collaborative model training for climate science, genomics, physics simulations, and experimental facilities.
Agents
Federated learning within agentic workflows and intelligent agents that orchestrate federated pipelines.
What technologies, metrics, and safeguards make federated AI successful in the long term?
03 / Program
The workshop combines invited talks with a panel discussion on current and future challenges. All the talks and presentations will be made available on this page following the conference proceedings.
04 / Speakers
Keynote and talk-title details will be added as the program is finalized.
Keynote
To be announced

Senior Scientist
NVIDIA, USA

Assistant Professor
University of Maryland, College Park, USA

Assistant Computational Mathematician
Argonne National Laboratory, USA

Associate Professor
Virginia Tech, USA
Invited speaker
To be announced
05 / Committee

Postdoctoral Research Associate
Oak Ridge National Laboratory, USA

Senior Research Scientist
Oak Ridge National Laboratory, USA

Associate Laboratory Director
Oak Ridge National Laboratory, USA

Distinguished Research Scientist & Group Leader
Oak Ridge National Laboratory, USA

Postdoctoral Research Associate
Oak Ridge National Laboratory, USA

Chief Senior Research Scientist
National Institute of Advanced Industrial Science and Technology, Japan

Senior Scientist & Group Leader
Argonne National Laboratory, USA

Senior Researcher
National Institute of Advanced Industrial Science and Technology, Japan

Computational Scientist & Group Leader
Oak Ridge National Laboratory, USA

Senior Scientist
NVIDIA, USA
06 / Previous workshops in the series
The International Conference for High Performance Computing, Networking, Storage, and Analysis
November 13-18 2022 · Dallas, Texas
The International Conference for High Performance Computing, Networking, Storage, and Analysis
November 12-17, 2023 · Denver, Colorado