Quantum computing, particularly annealing quantum systems, offers a practical near-term pathway to reduce AI's massive energy demands by solving certain computational problems far more efficiently than traditional processors. As artificial intelligence adoption accelerates across government and industry, the U.S. faces a critical energy bottleneck. The Electric Power Research Institute projects that AI data centers alone could consume up to 9% of U.S. electricity generation by 2030, raising serious concerns about energy shortages and higher electricity costs for consumers.

The timing challenge is acute. While demand for AI computing power is surging now, new power infrastructure can take a decade or more to permit, approve, and build. This fundamental mismatch between immediate demand and delayed supply is forcing companies to explore unconventional solutions, from building their own power plants to considering orbital data centers. Yet another option deserves serious consideration: leveraging quantum computing as a complementary technology to classical AI systems.

How Can Quantum Computing Reduce AI's Energy Footprint?

Quantum and AI are not competitors but complementary tools. AI excels at identifying patterns, generating predictions, and extracting insights from massive datasets, while quantum computing identifies optimal solutions to act on those insights. A breakthrough published in Science demonstrated the efficiency gap starkly: a problem was solved on a quantum computer in minutes using less than $1 of electricity. Solving the same problem on Oak Ridge National Laboratory's Frontier supercomputer would have taken nearly 1 million years and consumed more than the world's annual energy consumption.

The practical applications are already emerging. In drug discovery work with Shionogi, a Japanese pharmaceutical company, annealing quantum systems are showing promise in training generative models for novel molecular design faster and more efficiently. Forschungszentrum Jülich, one of Europe's leading scientific computing centers, has acquired an annealing quantum computer for integration with the Jülich UNified Infrastructure for Quantum Computing. This pairing is expected to be the world's first integration of an annealing quantum computer with an NVIDIA exascale supercomputer, providing a unique opportunity to observe how quantum and high-performance computing can work together.

What Real-World Applications Could Benefit From Quantum-AI Integration?

Beyond energy-efficient computing, quantum systems can help address complex challenges facing both public and private sectors. Unlike classical computers, which must evaluate options in sequence or rely on approximations for especially hard problems, quantum computers can explore many possible solutions simultaneously. This capability opens doors across multiple domains:

• Energy Grid Optimization: GE Vernova is investigating how quantum computing can identify vulnerabilities in the electric grid and optimize responses to potential disruptions or attacks, while E.ON, a European multinational electric utility company, has used annealing quantum computing to explore energy grid stability.
• Blockchain Efficiency: A quantum-assisted blockchain design could replace some energy-intensive classical proof-of-work with quantum-generated work, potentially reducing electricity use very substantially relative to classical-only approaches.
• Materials and Battery Design: Gate-model quantum computers, as they mature, are expected to be particularly powerful in designing better batteries and more efficient materials for energy storage and transmission.

These efforts demonstrate how quantum computing can strengthen the resilience and efficiency of critical infrastructure while simultaneously addressing the energy demands that AI is creating.

The integration of quantum computing into the administration's Genesis Mission, an ambitious effort that includes fusion, fission, materials, quantum, AI, advanced manufacturing, and grid modernization, could accelerate progress.

"Quantum computing may also result in efficiency improvements in other energy-intensive computing domains," noted Allison Schwartz, senior vice president of global government relations and public affairs at D-Wave.

Allison Schwartz, Senior Vice President of Global Government Relations and Public Affairs at D-Wave

Members of Congress have recognized the importance of both quantum modalities, annealing and gate-model, by including them in the National Quantum Initiative reauthorization bills currently moving through the legislative process. This legislative support signals growing recognition that quantum computing is not a distant future technology but a near-term tool for addressing current infrastructure constraints.

Why Should Policymakers Act on Quantum-AI Integration Now?

The expected benefits of combining quantum and AI are not decades away but rather could begin to materialize in practical applications in the near term. Relying solely on larger GPU clusters, more nuclear power plants, or infrastructure concepts like space-based data centers is not a sufficient strategy for managing AI's energy demands. These approaches should be combined with quantum computing, which is increasingly becoming a tool for addressing the compute, energy, and infrastructure constraints that AI is creating.

The challenge is substantial and immediate, but the solution space is broader than many policymakers currently recognize. If the goal is to scale AI without further harming grid reliability and affordability, leaders implementing the Genesis Mission and other infrastructure initiatives should recognize the impact that quantum computing could have today, not in the distant future.