The AI Supercomputer "Dossna" will change the landscape of scientific research... NERSC collaborates with Dell.

Artificial Intelligence (AI) and supercomputing technologies are revolutionizing the landscape of scientific research. Particularly noteworthy is the collaboration between the Lawrence Berkeley National Laboratory's NERSC (National Energy Research Scientific Computing Center) and Dell Technologies (DELL). The two parties are jointly developing a new generation supercomputer called the “Doudna” system, which is expected to be officially operational by the end of 2026, with preparations currently underway at a brisk pace.

Sudip Dosanjh, the director of the NERSC science division, emphasized: “A research community of over 11,000 people needs to work together for the deployment of this system,” and stated that to achieve a smooth transition of the system, close collaboration with Dell and NVIDIA is essential. The Dudenna system is a supercomputer based on the NVIDIA Vera Rubin platform and uses Dell's IR7000 racks. Considering efficient water resource management and sustainability, this system employs water cooling technology, and NERSC has also introduced dry cooling technology to reduce water consumption.

One of the core features of the system is its AI integration capability to perform real-time analysis of massive data and automate complex calculations. Tony Rea, the global AI infrastructure head at Dell, pointed out: “The most severe challenge in modern science is that the precision and scale of data are growing exponentially each year,” and explained that the Dudenna system will be a key tool in addressing this challenge.

Duo Sanghe revealed that the core of the design lies not only in AI computing power but also in the integration of complex workflows that support real-time feedback and experimental interaction. Researchers will be able to analyze data generated during the experiment instantly and adjust research directions in real-time based on the analysis results. This is expected to be a turning point that significantly enhances the efficiency of experiments in fields such as astronomy, physics, and life sciences.

The integration of AI and supercomputing is fundamentally changing the paradigm in the scientific community. Lei Ya stated: “As the threshold for scientific research lowers, every researcher can conduct deeper explorations at a faster pace,” emphasizing that the integration of AI has already shown results in both research productivity and efficiency.

The collaboration between NERSC and Dell goes far beyond system development. From data pipeline optimization to sustainable design, and to researcher-centric system architecture, this project is becoming an important model for demonstrating how AI technology can practically empower scientific research. This milestone will help the industry assess the commercial potential of future AI supercomputers and the upcoming transformation trends in the entire scientific community.