Guest post argues room temperature diamond qubits could be quantum computing's transistor moment

[RayOfLight87]

The Quantum Insider has a thought provoking guest post from Prof. Dr. Marius Grundmann, CEO of German company SAXON Q. His argument is that a recent experiment on a room temperature diamond quantum processor could be quantum computing's equivalent of the transistor replacing the vacuum tube. It is an opinion piece, and he runs a diamond quantum company, so keep that in mind. Still, the case he makes is interesting

The experiment he points to ran Grover's search algorithm on a commercial nitrogen vacancy diamond system, with no vacuum chamber or cryostat. It searched over eight possible states using three qubits. The calculation used long lived nuclear spins, which held their state for several milliseconds, with the electron spin acting only as an optical interface. The system found one marked state 77.3% of the time and two marked states 87.0% of the time, both above the best classical results for the same number of queries

Other numbers matter too. Average single qubit fidelity was 99.90%, and two qubit gate fidelities averaged 95.7%. The system ran at about 296 kelvin, which is room temperature, using roughly 600 watts. Dilution refrigerators for superconducting machines draw tens of kilowatts, so the power gap is huge, although Grundmann admits trapped ions still lead on raw fidelity

His historical point is that transistors did not beat vacuum tubes on performance at first. They won on manufacturability and a better improvement curve. He argues that superconducting, trapped ion and neutral atom machines all depend on vacuum or cryogenic equipment, which limits how far they can scale. His neatest line is that fabricating the qubit is not the bottleneck, housing it is

He also highlights a sulfur doping step that raises the yield of usable centres to around 85%, compared with about 1% for older methods. He thinks room temperature, low power qubits could open markets in vehicles, robots, edge devices and hospitals. He even suggests the billions invested in cryogenic approaches could face the same fate as CRT screens. That is a bold claim, but it is the kind of argument worth debating

Could diamond qubits overtake the cryogenic approaches? And is the transistor comparison fair?

Guest Post: Quantum Computing's Transistor Momentthequantuminsider.com

[IronFist38]

Melbourne opened a quantum diamond hub recently, and Japan has a room temperature machine running too. Diamond and other room temperature approaches are getting more attention

It would be a big shake up if one of them broke through. Exciting to watch

[KyleOReilly]

The transistor comparison is clever but stretched. Transistors were better in almost every way within a decade. Diamond qubits still have much lower two qubit fidelities than the best ions and superconducting chips

They need to close that gap before the comparison holds. Ninety five percent two qubit fidelity is far from fault tolerance

[Jedi Stuart]

Sulfur doping lifting yield from 1% to 85% is the most important number in the piece. Yield is everything in manufacturing

If you can reliably make good centres, you can start thinking about scale. That is the real story. That alone could make diamond devices far cheaper

[OffsideTrap66]

Keep in mind who wrote it. A CEO of a diamond quantum company arguing that diamond is the future is not exactly neutral

That does not make him wrong. But I would like to hear from independent experts too. Guest posts are useful for ideas, less so for evidence

[Badger90]

Fabricating the qubit is not the bottleneck, housing it is. That line sticks with me. Anyone who has seen a dilution fridge knows how much engineering surrounds a few qubits

Getting rid of that would be transformative. The fridge is often bigger than the computer it serves