RIKEN keeps five silicon spin qubits above 99.9% fidelity when driven together

Started by NightOwl83, Yesterday at 07:35 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Topic: RIKEN keeps five silicon spin qubits above 99.9% fidelity when driven together   Views(Read 72 times)
Active members in this topic:
NightOwl83(1) Brooke_40(1)

NightOwl83

A team from RIKEN, National Taiwan University, ETH Zurich and TU Delft has solved a stubborn problem with silicon spin qubits. Quantum Computing Report's Mohamed Abdel-Kareem reports that the researchers ran high fidelity single qubit gates on five qubits at the same time, all through one shared microwave line. The work is published in Nature Physics

Driving qubits in parallel has always come with a cost. When several qubits were controlled together, fidelity used to drop from above 99.9 percent to around 99.5 percent. That might sound like a small difference, but for error correction it is a big one

The team traced the problem to AC Stark shifts, where the drive signal meant for one qubit slightly shifts the frequency of its neighbours. They fixed it with a pairwise phase calibration that accounts for every pair of qubits. With that in place, individual gates reached between 99.993 and 99.998 percent, and two or three qubits driven together stayed above 99.99 percent

Driving all five at once still gave at least 99.90 percent. The device was a five qubit quantum dot array made from isotopically purified silicon 28 and silicon germanium. The authors suggest modules of around ten qubits sharing a single control line could be a practical design

This matters because wiring is one of the biggest obstacles to scaling. Every qubit needing its own control line quickly becomes impossible in a fridge. Sharing lines without losing fidelity is exactly the kind of engineering progress silicon qubits need

Is silicon the most promising route to large scale quantum computers? And how big a problem is wiring for scaling?


Brooke_40

Sharing one control line across several qubits is a big deal. Cryostats only have so much room for cables. Fewer lines means more qubits in the same fridge. Dilution fridges have strict limits on heat load. Each cable adds a little warmth