Google DeepMind and Commonwealth Fusion Systems (CFS) formed a partnership that will leverage AI to accelerate a path to a viable fusion power grid. CFS is a fusion power company currently developing a nuclear fusion power plant in Virginia. Fusion power offers clean, virtually limitless energy. It’s the same process that powers the Sun. However, scientists, researchers, and engineers have yet to find a way to perfect the process.
By combining AI and fusion technology, DeepMind and CFS hope to make commercial fusion energy a reality.
Unlocking Limitless Energy With AI
Fusion requires the containment of plasma at a blazing temperature of over 100 million degrees Celsius within a tokamak. The tokamak is a magnetic confinement machine. Keeping this extremely volatile state is immensely challenging. However, CFS’s compact and powerful tokamak, SPARC, aims to be the first magnetic fusion machine to achieve “breakeven.” This achievement would generate more power from fusion than it takes to sustain it.
Explore Tomorrow's World From Your Inbox
Get the latest science, technology, and sustainability content delivered to your inbox.
I understand that by providing my email address, I agree to receive emails from Tomorrow's World Today. I understand that I may opt out of receiving such communications at any time.
The recent partnership focuses on integrating DeepMind’s advanced AI directly into the CFS workflow across three critical areas: high-speed simulation, operational optimization, and real-time control.
Google DeepMind’s open-source plasma simulator, TORAX, plays a central role in this AI and fusion collaboration. According to Google, TORAX enables CFS to run millions of virtual experiments to test SPARC’s operating plans before even powering on the machine. Devon Battaglia, Senior Manager of Physics Operations at CFS, said, “TORAX is a professional, open-source plasma simulator that saved us countless hours in setting up and running our simulation environments for SPARC.”
The AI’s strength is in rapidly finding a path to maximum energy. Reinforcement learning and evolutionary search approaches allow AI agents to explore vast numbers of potential operating scenarios in simulation. Utilizing this, CFS can pinpoint robust and efficient strategies to increase the probability of success when SPARC is fully operational.
By introducing AI to discover complex adaptive strategies, the partnership aims to accelerate SPARC’s progress toward its goal of bringing the world closer to a future powered by fusion energy.
