The Quantum Dragon had a nightmare about trying to rescue people from drowning. At a minimum, he had to help them keep their heads above the water. What could that possibly mean?
Before he went to bed in his cave, he had just finished reading “Real-Time Adaptive Tracking of Fluctuating Relaxation Rates in Superconducting Qubits.” He had found out about it through “How to improve the performance of qubits: super-fast fluctuation detection achieved at NBI” and “Listening to the Qubit’s Heartbeat: real-time tracking of T₁.”
What we came up with is that Niels Bohr Institute and Quantum Machines have achieved real-time calibration because you can’t just characterize your qubits once. You can do your best with the qubits that you have identified as “good” and “bad,” but the definition of “good” and “bad” changes in milliseconds. This is surely a nightmare for everyone involved, so we assume this is why he had a nightmare in the first place.
The breakthrough is quick measurements over a span of seconds. The hardware gets useful information in seconds instead of hours, allowing estimation of decay rates based on physics. Instead of using averages, there are now more data points for statistics.
So, what’s the impact? This is very low level, so it’s hard to quantify the impact on algorithms. Quantum error correction (QEC) is still needed. But QEC is limited by the worst qubit, and therein lies the tie-in. All the qubits have to stay above a certain threshold for QEC to be possible, and that’s what this protocol helps to do. Therefore, The Quantum Dragon’s nightmare about helping people keep their heads above water was really about helping to keep qubits above threshold for QEC.
A side benefit of using this protocol is that the statistics tell you whether the fidelity is good enough to execute your algorithm. This gives you the option to wait, and although waiting can be slow, it avoids wasteful computation. And if you’re paying for access, this can save you quite a bit of money.
