Skip to 0 minutes and 4 secondsI am here again at the University of Tokyo with Professor Yasunobu Nakamura, an expert on superconducting qubit. Professor Nakamura, tell us about devices you are building. Okay, we are trying to build a small quantum information processing system using superconducting qubits. So there we put several superconducting qubits as an array and then control them to manipulate the quantum state to demonstrate simple quantum information processing. How do you execute one and two qubit gates using superconducting system?
Skip to 0 minutes and 40 secondsYes, so as you know, it is very important to apply very precise control on superconducting qubit and for that we are using microwave pulses typically in a timescale of 10 nanosecond to 100 nanosecond to implement quantum gate for single qubit control and 2 qubit gate. How do you make your superconducting chip? Yes, similarly to silicon technology, we use lithography for making superconducting devices, so like optical lithography and electron beam lithography are used to create superconducting chips in a tiny scale. Okay, what is unique about your approach?
Skip to 1 minute and 28 secondsOkay, so now we would like to put superconducting qubits in array in a two-dimensional way, so qubits will be arranged on a small chip in a two-dimensional array and then we would like to control the quantum state to implement simple quantum information processing. Okay. Tell us what's the challenge for building large scale quantum system using your technique.
Skip to 1 minute and 58 secondsOkay, the important thing is that we need to control many qubit, so for that we need many wiring to control and read out the signal and that is quite challenging because chip – on the chip qubits are densely aligned and then there are not much – there is not much space left for the wiring, so the arranging all the wiring properly and also sending and read out the signal properly is quite challenging. That’s our near term target. Okay, thank you professor Nakamura.
In an earlier Step, we visited the laboratory of Professor Yasunobu Nakamura at the University of Tokyo, and learned about using electric charge and magnetic flux as our qubit state variable. In this video, we continue that visit and learn more about how those experiments are conducted.
Experiments in a refrigerator
The large can behind Professor Nakamura in the video is a dilution refrigerator, which has several stages that are each progressively cooler. Inside the dil fridge is a large rig that passes through the stages, like this:
The figure below shows how a qubit is controlled from the outside (as we saw in the video), with different parts of the circuit at different temperatures: room temperature (“RT” in the figure), 4 Kelvin, then 10 millkelvin – 1/100th of a degree above absolute zero! (Quantum dot experiments use a similar physical setup to cool their chips.)
At the bottom stage is a block of aluminum with two halves, each with a single connector for a coaxial cable. Mounted inside is a small, rectangular chip, a few millimeters in size. This chip has a single transmon qubit that is about one micron in size, as you can see in the figure below. The Josephson junction itself is small, but including the resonators necessary to shield and control the qubit, the overall size is nearly a millimeter.
The figure below is a computer rendering of a new chip design. The qubits are the yellow concentric circles at each lattice site. The inner and outer rings are bridged with a Josephson junction to form a transmon qubit. The qubit has four arms to interact with its neighbors. The qubits are controlled from below the chip, as you can see in the cutaway.
The four meander lines are readout resonators, used to measure the state of a qubit. Four qubits share one readout connection which is connected to a coaxial cable beneath the chip (not shown).
The small white holes that pass through the chip connect the red sheets on the top and bottom, called ground planes, to stabilize the voltage used as ground. The yellow control lines from the bottom are coaxial cables, shown in cutaway, used to execute one-qubit rotations and couple two neighboring qubits, so that we can perform two-qubit gates like CNOT.
The picture is just a schematic and not the final design, but it gives you the idea of what it takes to create a system, rather than just a single qubit in an experiment. If this design is successful, Professor Nakamura’s team will be able to build chips with tens or hundreds of qubits on them. Designs like this may be able to run surface code error correction, which we will introduce in the next Activity.
Superconducting qubits, especially transmons, are large size compared to transistors. We won’t be able to put many qubits into a single chip, and entangling two superconducting qubits in separate chips is one of the major outstanding challenges.
© Keio University