Quantum physics for quantum physicists. Praising the cool stuff, snarky comments about the rest. Personal account: @ekwilibrator. Opinions my own!
Joined February 2019
- Tweets877
- Following603
- Followers1.3K
- Likes867
The sweet thing about a really snowy day is that people somehow start saying "hello" to each other when crossing on the street
A similar method allows us to tune two-qubit gates on a per-zone basis. We measured the Bell states in the central zone with high precision (though all zones seem identical for all we know), recording a record 99.97(1)% fidelity, and impressive stability over 60 hours. 11/n
Not only is this method chip-integrated, it also works *really* well. In the paper, we perform 1-qubit gates in any zone with < 1e-5 error and ~ 1e-6 level crosstalk. And it's really scalable - for all we know, we can just keep adding zones and they will all work the same! 10/n
which allows us to make the Rabi flops either very slow (qubit 'hidden') or very fast (qubit 'active'). We can do it very flexibly on a per-zone basis, allowing us to e.g. make an X gate in any subset of zones in the processor. 9/n
Of course, it's not enough to do gates everywhere, we also need to modulate them locally. But this can be done electrically by fine-tuning the ion position above the current-carrying trace. In our experiment, we apply voltages applied to DC electrodes to move the ion along y 8/n
What's neat about this method is that you can drive a lot of qubits from one trace, i.e. instead of building another laser lab every time you want to add a few qubits, all you need to do is make the on-chip trace a little longer. In the end, your QC looks simply like this 7/n
At @OxfordIonics we decided to follow a different, more scalable paradigm. Our ions are also trapped above a chip, but quantum gates are driven not by lasers, but by AC currents running through chip traces. For background see e.g. nature.com/articles/nature10… 6/n
Case and point: arguably the most powerful QC in the world, @QuantinuumQC's H2, has 56 qubits but only 4 gate zones (circled) that the qubits are sequentially placed into. This is because scaling free-space optics is a PITA 4/n
On the other hand, trapped-ion qubits have really low noise: >99.9% fidelity, >1h coherence time etc. But all the commercial trapped-ion systems rely on free-space optics to deliver high-power control beams, which is really tough to scale... 3/n
