Quantum Computer Energy Use vs Supercomputers: What Is Actually Known
The claim you hear, and the reality
You will see two opposite stories. One says quantum computers will be vastly more energy-efficient than supercomputers. The other says they are power-hungry fridges. The honest status is that nobody can fully settle it yet, because useful, large quantum computers do not exist and vendors rarely release power data. This guide sticks to what a careful 2026 report from Physics Today states.
What is reported
The Physics Today article (June 2026) is notable for what it does not give: it lists no kilowatt figure for any specific quantum computer. Its numbers are projections and estimates:
- A hypothetical code-breaking machine. A 2023 RAND analysis by Parker, as summarized, estimated that a superconducting computer with 20 million noisy qubits cracking a 2048-bit RSA key would need about 890 megawatt hours, roughly a year of electricity for 85 US households, costing about 64,000 dollars in 2023 dollars. It gives no runtime, so no average power. See Shor's algorithm.
- Six designs scaled to 4,000 logical qubits. A 2025 presentation by Olivier Ezratty put base power from under 1 megawatt to over 100 megawatts. The article stresses this was not peer reviewed and was for demonstration, and that most vendors do not share power data. It says most projected figures fall within range of the top supercomputers today.
- Oak Ridge projection. The article says a McCollum et al. analysis found integrated quantum-classical data centers using superconducting qubits would, by the 2040s, have power demand of the same order of magnitude as today's data centers.
Where the electricity goes
For a noisy intermediate-scale superconducting machine, the article cites a May 2026 preprint finding that cooling accounted for approximately 80 percent of power consumption. Neutral-atom machines need high-powered lasers and trapped-ion machines need cooling and lasers, but the article gives no figures. A Quandela preprint from 2026 reportedly found its photonic computer more energy efficient than a classical computer on one algorithm, despite a longer runtime. That is one algorithm, not a general result. Projected power at larger scale may shift toward control electronics.
What about supercomputers?
The article does not cite wattage for specific supercomputers. For context it notes that data centers used 1.5 percent of world electricity in 2024 according to the IEA, with consumption expected to more than double by 2030, and that research supercomputers use far less than commercial data centers. We do not repeat unsourced per-machine megawatt figures here.
Why a fair comparison is hard
- Today's quantum machines do not reliably solve useful practical problems, so energy per useful answer cannot be measured.
- Efficiency claims depend on specific algorithms and on whether a quantum advantage exists for that task (see quantum advantage).
- Claims from companies in the field should be weighed accordingly.
- If quantum stays in specialized research, total impact could be small; if costs fall, the article notes a rebound effect could raise total use.
Hybrid reality
Real deployments pair quantum chips with classical GPUs for error decoding and control (see hybrid GPU quantum systems and error correction). That means the classical energy bill belongs in any honest accounting, and published figures rarely include it.
How to read an energy claim
When you see a headline about quantum energy savings or waste, ask five plain questions. Is it a measured machine or a projection? Does it count cooling, control electronics and the classical computers beside the quantum chip? Which task is being compared, and has a quantum advantage actually been shown for it? Is the author a company or lab with something to sell? Was it peer reviewed? The Physics Today report itself notes that the Ezratty calculation was not peer reviewed and that vendors rarely publish power data, and that is exactly the kind of caveat worth looking for. Where an answer is missing, the right reading is "unknown," not "good" or "bad."
Also keep scale in mind. A research quantum computer is one machine in one lab. Data centers, by contrast, are measured as a share of national grids. The comparison that matters in future is not machine against machine but total energy per useful answer, and that number does not exist yet for any quantum workload.
The hopeful angle
The point is not that quantum is green or wasteful; it is that this is an open engineering target. Cooling dominating the budget suggests clear places to improve, such as better fridges, higher-temperature qubits and efficient control chips. And researchers are asking the question early, which is healthy. Not financial advice, and nothing here predicts any asset price including QNT, an independent memecoin with no link to Quantinuum Ltd.
Sources and further reading
Reported as of 2026-10-09. Supply deals, prices and rules change often, so check the primary documents. Nothing here is financial advice. The QNT memecoin is independent of Quantinuum Ltd, the real company, and of every supplier, lab and government named on this page.
Frequently asked questions
How much power does a quantum computer use?
Public data is thin. The Physics Today report we relied on gives no kilowatt figure for any specific machine, and says most vendors do not share power data.
Do quantum computers use less energy than supercomputers?
Not established. Projections for scaled machines range from under 1 megawatt to over 100 megawatts, and today's machines cannot yet do useful work to compare per answer.
Where does a superconducting quantum computer's power go?
A May 2026 preprint, as reported by Physics Today, found cooling accounted for approximately 80 percent of one noisy superconducting machine's power.
Keep reading
- How Are Quantum Computers Cooled?
Why do some quantum computers sit in huge fridges near absolute zero? A plain English guide to dilution refrigerators, noise and which qubits need cooling. - Hybrid GPU and Quantum Systems in Practice: The Latency Budget
A closer look at why quantum computers need GPUs next to them: real time decoding, calibration, microsecond links, and the Quantinuum Helios decoding demonstration. - Quantum Supremacy vs Quantum Advantage Explained
What the terms quantum supremacy and quantum advantage mean, and why headline claims are often debated.
All Quantum industry, people and AI guides | Back to top | Search the site
Main pages: Quantum computing explained | Quantum and crypto | Companies | Quantum news | Glossary