Google и Commonwealth Fusion Systems создают ИИ для управления термоядерным синтезом Translation: Google and Commonwealth Fusion Systems Create AI for Controlling Nuclear Fusion

Google Corporation is set to develop artificial intelligence aimed at overseeing the process of nuclear fusion. To achieve this, the company has partnered with Commonwealth Fusion Systems (CFS).

The collaborative initiative focuses on the creation of AI agents designed to monitor plasma in the SPARC reactor and enhance its operation. The ultimate goal is to draw closer to the establishment of a nearly limitless energy source.

«Nuclear fusion—the process that powers the Sun—promises humanity clean and virtually inexhaustible energy without long-lived radioactive waste,» states Google’s press release.

The partners are concentrating on three primary areas of research. In the first area, Google and CFS have developed an accurate open-source simulator called TORAX, enabling virtual experiments with various operational modes of the facility.

This system is built on the JAX platform and can function on both CPU and GPU, integrating artificial intelligence models to boost performance.

“[TORAX] has saved us countless hours in setting up and running models for the SPARC project,” noted Devon Battaglia, Senior Manager of Physical Operations at CFS.

The second focus involves employing reinforcement learning to identify optimal parameters for the nuclear reactor’s operation. The AI agents analyze millions of potential configurations of magnetic coils, fuel input systems, and heating power to achieve a positive energy balance.

The third focus is on developing real-time systems to manage thermal loads. SPARC will generate intense heat concentrated in small areas, and the AI must learn to dynamically distribute these loads.

Additionally, Google has become an investor in CFS, supporting the commercialization of the technology. The SPARC reactor is anticipated to become the first magnetic fusion facility to surpass the breakeven energy threshold, where the generated power exceeds the costs of sustaining the reaction.

It is worth noting that in October, the company collaborated with Yale University to develop an AI model for cancer treatment.