STELLAR-AI Platform Launched to Accelerate Fusion Energy Research

STELLAR-AI: A Leap Forward in Fusion Energy Research

The U.S. Department of Energy’s Princeton Plasma Physics Laboratory (PPPL) has launched the STELLAR-AI platform, a groundbreaking initiative designed to accelerate fusion energy research by integrating artificial intelligence (AI) with high-performance computing. This innovative platform aims to overcome the bottlenecks in fusion research by speeding up simulations and experiments, bringing the world closer to commercially viable fusion power.

The STELLAR-AI project, led by PPPL, will unite national laboratories, universities, technology companies, and industry partners to create a computational foundation tailored to the needs of the fusion community. By combining CPUs, GPUs, and QPUs in an optimal hardware configuration, STELLAR-AI will tackle the complex challenges facing private fusion companies as they race to develop sustainable energy solutions.

Integrating AI with Fusion: The STELLAR-AI Approach

At the heart of STELLAR-AI is the integration of AI with experimental devices, such as PPPL’s National Spherical Torus Experiment-Upgrade (NSTX-U), which is set to go live this year. This integration will enable researchers to analyze data in real-time as experiments occur, significantly reducing the time required to run high-fidelity simulations or train AI systems capable of designing optimal fusion systems.

Jonathan Menard, deputy director for research at PPPL, emphasizes the platform’s potential: "Fusion is a complex system of systems. We need AI and high-performance computing to optimize the design for economic construction and operation." By linking simulation technology with experiments, particularly NSTX-U, and leveraging AI and partnerships, STELLAR-AI aims to accelerate the path to commercially viable fusion energy.

A Cornerstone of the U.S. Fusion Ecosystem

STELLAR-AI is part of the Genesis Mission, a national initiative launched by executive order in November 2025 to accelerate scientific discovery across DOE laboratories using AI. The platform will contribute fusion-specific computer codes, data, and scientific models back into the national system, aligning with the DOE’s Fusion Science and Technology Roadmap.

The project will support various applications, including creating a digital twin of NSTX-U, a virtual model that mirrors the physical machine for testing ideas before actual experiments. Another key project, StellFoundry, will use AI to expedite the design of stellarators, a type of fusion device with a twisted, pretzel-like shape that some scientists believe offers advantages over other designs.

With partnerships spanning DOE National Laboratories, AI and HPC companies, academic institutions, and fusion industry leaders, STELLAR-AI is poised to deliver proven AI models and key tools for the U.S. fusion industry. This collaborative effort positions PPPL as a hub for public-private collaboration in fusion energy, leveraging its seven decades of plasma research and computational expertise.