At the SCSP AI+ Expo, U.S. Energy Secretary Chris Wright and NVIDIA Vice President Ian Buck shared a compelling vision for how artificial intelligence can help build the energy infrastructure needed for the next era of American leadership. In a fireside chat moderated by SCSP President Ylli Bajraktari, they argued that the nation's dominance in AI depends on its ability to produce abundant, affordable energy. Their discussion centered on the U.S. Department of Energy’s (DOE) Genesis Mission, a bold initiative to apply AI to scientific discovery, and NVIDIA’s pivotal role in powering that mission. Below are key insights from the conversation.
What is the core argument linking AI and energy leadership?
Secretary Wright and Ian Buck made a straightforward case: American leadership in artificial intelligence cannot happen without American leadership in energy production. Wright emphasized that energy is the foundation of life and opportunity, stating, “The more energy you have, the more affordable energy you have, the more opportunities you have in your society.” AI systems, particularly large-scale models, require enormous amounts of electricity. By advancing energy technologies—through nuclear, renewables, and next-generation sources—the U.S. can fuel both its AI ambitions and its economic growth. Buck added that NVIDIA’s investments in energy-efficient computing are designed to support this symbiotic relationship. The two leaders agreed that the path to technological supremacy runs through reliable, low-cost power, making energy policy a critical pillar of national strategy.
What is the Genesis Mission and how does it connect AI to scientific discovery?
The Genesis Mission is the DOE’s flagship effort to harness artificial intelligence for accelerating breakthroughs in science and energy. By combining the agency’s 17 national labs, world-class scientists, and vast datasets, the mission aims to solve some of the toughest challenges—from climate modeling to materials science. AI tools developed under Genesis will help researchers analyze data faster, simulate complex systems, and uncover new insights that would otherwise take years. Secretary Wright described Genesis as a practical demonstration of how AI can directly benefit the energy sector. Buck called it “the most exciting initiative” he has seen across labs and industry, noting that NVIDIA is fully committed to providing the necessary computing power and software. The mission essentially turns AI into a force multiplier for national laboratories, enabling them to tackle problems at an unprecedented scale.
How is NVIDIA partnering with the DOE on the Genesis Mission?
NVIDIA’s partnership with the DOE goes back two decades, and the Genesis Mission deepens this collaboration. Ian Buck stressed that NVIDIA brings more than just hardware—it offers a complete computing stack, including algorithms, software frameworks, and decades of expertise in building supercomputers for scientific use. The company is co-developing two major AI supercomputers at Argonne National Laboratory. The first, Equinox, is currently being deployed with 10,000 NVIDIA Grace Blackwell GPUs—the same chips and software used by leading AI labs worldwide. The second, Solstice, will feature 100,000 next-generation NVIDIA Vera Rubin GPUs. Buck explained that this scale of computing power, dedicated exclusively to science, represents a monumental leap. He described it as “five times larger than the entire TOP500 supercomputer list combined,” measured in exaflops. This infrastructure will enable researchers across the globe to access cutting-edge AI capabilities for scientific discovery.
What are the specifications of the Equinox and Solstice supercomputers?
Equinox, nearing completion at Argonne, is powered by 10,000 NVIDIA Grace Blackwell GPUs. These are the same advanced processors used in today’s leading AI training systems, ensuring that scientific researchers benefit from the latest innovations. Solstice, the follow-up system, will be significantly larger: it will incorporate 100,000 NVIDIA Vera Rubin GPUs, the next-generation architecture designed specifically for high-performance computing. According to Ian Buck, Solstice will deliver approximately 5,000 exaflops of computational performance. To put that in perspective, that is five times the combined power of the entire TOP500 list of the world’s fastest supercomputers. Both systems are built using the same hardware and software building blocks that power major AI labs, making them accessible for a wide range of scientific applications. This unprecedented computing capacity will enable researchers to simulate complex physical phenomena, accelerate materials discovery, and optimize energy systems at a scale never before possible.
How does NVIDIA’s open-source AI model trained on physics papers help?
As part of its commitment to the Genesis Mission, NVIDIA developed an open-source AI model trained on a massive corpus of scientific literature. The model was initially trained on 1.5 million physics papers, then fine-tuned on 100,000 papers specifically focused on energy and materials science. This specialized model understands the language and concepts of physics, making it a powerful tool for researchers. It can help scientists quickly search through decades of publications, generate hypotheses, and even suggest new experimental approaches. By releasing the model as open source, NVIDIA ensures that the entire scientific community can benefit, not just those at top-tier institutions. Ian Buck emphasized that this approach democratizes access to AI, allowing any lab or university to leverage state-of-the-art machine learning for their research. The model is part of a broader strategy to make AI a common resource for accelerating scientific discovery, particularly in fields critical to energy innovation.
Why does Secretary Wright consider affordable energy essential for societal opportunity?
Secretary Chris Wright framed energy as the bedrock of modern civilization. He argued that affordability and abundance of energy directly correlate with the opportunities available to people. When energy is cheap and plentiful, societies can power homes, schools, hospitals, and factories, raising living standards and enabling economic growth. In contrast, expensive or scarce energy limits development and exacerbates inequality. Wright linked this directly to AI: to build and run the data centers that train large models, the U.S. needs reliable, low-cost electricity. By investing in diverse energy sources—including nuclear, renewables, and natural gas—and advancing technologies like next-generation batteries and grid modernization, the country can secure its energy future. This, in turn, supports AI leadership, which itself can optimize energy systems for even greater efficiency. The message was clear: energy and AI are not competing priorities; they are mutually reinforcing pillars of American prosperity and global competitiveness.