Quantum computers may offer a surprising solution to artificial intelligence's escalating power demands, not by replacing GPUs entirely, but by handling a specific, energy-intensive task that classical computers struggle with: generating high-quality training data. Australian startup Diraq argues that utility-scale quantum systems could perform complex simulations far more efficiently than traditional processors, potentially cutting the energy footprint of AI model training.

Why Is AI Training Data Generation So Energy-Intensive?

The creation of high-quality training data represents a largely overlooked driver of artificial intelligence's energy consumption. Classical computers struggle with complex simulations needed to generate robust datasets for tasks like molecular design and power grid optimization. This limitation forces AI developers to rely on energy-intensive classical compute to generate approximations, consuming enormous amounts of electricity in the process.

Some new artificial intelligence models are expected to exceed 1 megawatt of power consumption, using more electricity than an average US home consumes in two years. Global data center electricity generation is projected to surge toward 1,000 terawatt-hours by 2030, according to Diraq's analysis.

How Could Quantum Computing Help?

Diraq proposes that utility-scale quantum computers could perform these simulations efficiently, effectively replacing the racks of GPUs currently used for approximate training data generation. This approach differs from more publicized quantum applications like molecule design by focusing on the upstream problem of data creation rather than downstream problem solving.

The company's strategy centers on scaling qubit density on silicon chips fabricated using standard CMOS foundry lines, aiming for millions of qubits at a cost of under a dollar per qubit. Founded in 2022 and headquartered in Sydney, Australia, Diraq has raised over $140 million to pursue this silicon-spin qubit technology.

A September 2025 Nature paper demonstrated above 99 percent single- and two-qubit fidelity on industry-fabricated silicon unit cells, surpassing the surface-code fault-tolerance threshold and validating the potential of this approach, according to the company.

What Are Diraq's Power Efficiency Targets?

Diraq is targeting a 300 kilowatt total system power draw for its utility-scale quantum computer, a figure intended to align with the operational constraints of modern data centers. This target accounts for both the quantum processing unit itself, expected to consume around 125 kilowatts, and the classical compute infrastructure required for interfacing and control.

Currently, Diraq has deployed an eight-qubit system in a commercial data center, drawing under 20 kilowatts for the complete quantum system. The system's compact design includes self-contained cryogenic cooling and control electronics, allowing deployment alongside conventional computing equipment without requiring a dedicated quantum facility.

How to Understand Quantum's Role in AI Infrastructure

• Complementary Technology: Quantum computers will work alongside classical systems, with GPUs and other processors handling orchestration, error correction, AI workloads, and computations quantum machines aren't designed to perform.
• Scalability Timeline: Diraq's roadmap projects thousands of physical qubits in a commercial product by 2029, and tens of millions by 2033, representing a gradual path to utility-scale deployment.
• Infrastructure Integration: Silicon spin qubits minimize the need to scale peripheral infrastructure, contrasting sharply with other quantum architectures like superconducting and photonic systems, which require expansive cryogenic or optical plants.
• Cost Efficiency: The company argues that increasing on-chip qubit density is the primary constraint, not qubit quality, and that avoiding the construction of new dedicated quantum facilities eliminates significant embedded emissions and costs.

"Quantum computing will ultimately be judged by the problems it solves, but it will be adopted according to the economics of delivering those solutions," Diraq stated, highlighting the importance of practical considerations alongside scientific advancements.

Diraq, Company Statement

The company is currently integrating its quantum processor with a Dell high-performance computing cluster for low-latency hybrid workflows, further demonstrating its commitment to seamless integration with conventional computing resources.

What Does This Mean for the AI Infrastructure Race?

The broader context matters here. While quantum computing offers potential energy savings for specific tasks, the AI infrastructure race is increasingly about assembling the financing, energy, hardware, and data center capacity needed to deploy AI at scale. Major cloud providers are investing heavily in inference capacity, with combined capital expenditures of nine major global cloud providers expected to exceed $886.7 billion this year.

The strategic advantage in AI infrastructure may belong not solely to whoever manufactures the fastest processor, but to those capable of connecting customers with the capital, power, and compute necessary to turn AI ambition into operating infrastructure. Diraq's approach of integrating quantum systems into existing data center environments positions the technology as a practical tool for reducing energy consumption in specific workflows, rather than a wholesale replacement for classical computing.

"Quantum does not automatically mean energy efficient," Diraq noted, emphasizing that every quantum architecture must reach millions of physical qubits, but the path to that scale varies considerably. The company's focus on density and practical integration suggests that quantum computing's role in addressing AI's energy demands will be measured and incremental, not revolutionary.