Magne: The Danish Quantum Computer Challenging Energy

Magne: A Milestone for Europe

Europe is set to take a significant step in the field of quantum computing with the introduction of Magne, touted as the most powerful quantum computer on the continent. Scheduled to be operational in Denmark later this year, this ambitious project is the result of a collaboration between Microsoft and Atom Computing. It has received substantial financial backing of 80 million euros from the Novo Nordisk Foundation and the EIFO. While this advancement is promising, it raises crucial questions about the future of quantum computing in Europe, particularly regarding large-scale deployment.

The Challenges of Neutral Atoms

Atom Computing's platform relies on the use of neutral atoms, an approach that offers distinct advantages, including the avoidance of the dilution cooling required for superconducting systems. However, this method still necessitates a complex infrastructure, including vacuum equipment, lasers, optical tweezers, detectors, and sophisticated electronic systems. Although the qubits are compact, the overall infrastructure remains bulky, complicating large-scale expansion.

The Potential Impact of Quantum Computing

As quantum computing becomes commercially viable, it could transform various sectors, from drug discovery to aerospace. However, the industry will need to meet a growing demand for millions of quantum computers. This prospect poses challenges in terms of size, weight, power, and unit costs—crucial factors for the future development of this technology. There is often talk of millions of qubits as a utility threshold, but it is equally important to consider what will happen when the market demands millions of these machines.

Alarming Energy Forecasts

According to BloombergNEF, the energy demand from data centers in the United States could reach 106 gigawatts by 2035, a 36% increase from previous forecasts. This significant rise underscores the urgency of planning for the integration of quantum computing, whose energy needs could surpass those of current AI. Size, weight, power, and unit economics will be just as important in quantum computing as they are in AI.

Energy Requirements of Quantum Systems

An independent economic study by ICM compared the energy needs of different approaches to quantum computing. To achieve 4,000 logical qubits, superconducting, photonic, and ion trap systems require approximately 160, 100, and 140 megawatts, respectively. These figures are comparable to those of modern hyperscale AI data centers, illustrating the scale of infrastructure needed to support these machines. These are not modest additions to network demand, but entire facilities built to serve a single machine.

The Massive Infrastructures of PsiQuantum

PsiQuantum, a company backed by $940 million in Australian government funding, is building a 540,000 square foot facility near Brisbane Airport to house its quantum computing modules. This massive infrastructure includes a cryogenic center and a main office building, highlighting the logistical and energy challenges associated with large-scale quantum computing. It illustrates what large-scale quantum computing looks like according to these approaches, and it relies on the same networks already under pressure from AI.

Quantum Motion's Innovative Approach

Quantum Motion is focusing on silicon spin qubits, manufactured on 300 mm CMOS production lines that produce chips for laptops and smartphones. This approach promises a significant reduction in energy consumption, with a target system requiring less than 200 kW—about a thousand times less than other modalities at an equivalent computational scale. However, engineering challenges remain, particularly in cryogenic control and uniform manufacturing.

Promising Advances in Engineering

Recent advancements, such as cryo-CMOS control at millikelvin temperatures and the integration of 1,024 devices under 1K, indicate that current obstacles are primarily technical. These developments pave the way for more compact and energy-efficient quantum computers capable of integrating into existing data centers. Recent results suggest that these are engineering problems rather than fundamental ones.

Anticipating the Energy Impact of Quantum

The rise of AI has already surprised many policymakers with its energy impact. Quantum computing could pose similar, if not greater, challenges. It is crucial to assess the future energy needs of these technologies now to avoid overloading already strained networks. Europe, with its ambitious projects like Magne, must anticipate these issues to ensure a smooth transition into the quantum era. The question of how much energy a useful quantum computer will consume, and where it will be used, deserves the same scrutiny applied to AI infrastructure.