The Nature sub-issue reports a new microwave source that could facilitate the development of large-scale quantum computers

Quantum computing is the current research hotspot, and scientists are working hard to expand the scale of quantum computing and develop practical quantum computers. Recently, Nature-Electronics reported a circuit that produces the high-quality microwave signals needed to control quantum computers near absolute zero, which could greatly increase the number of qubits in a quantum processor.

The study was done by researchers led by Aalto University and the VTT Technology Research Center in Finland, and the paper was published on December 9 local time under the title "A low-noise on-chip coherent microwave source" (a low-noise chip coherent microwave source).

"Our work serves large-scale quantum computers, especially superconducting quantum computers. Current superconducting quantum computers, while demonstrating superiority over specific tasks, are still far from large-scale general-purpose quantum computers. Yan Chengyu, the first author of the paper and a postdoctoral fellow at Aalto University (now an associate professor at the School of Physics of Huazhong University of Science and Technology), told the surging news (www.thepaper.cn) reporter.

Yan Chengyu introduced that the realization of large-scale quantum computers requires two core technologies: First, find an architecture that realizes more qubit coupling. Second, realize the chip integration of qubits and control modules.

"The first point is done by a lot of people, and the second point gets less attention, but it is really a bottleneck." Yan Chengyu said this is due to the architecture of existing superconducting quantum computers: the quantum processor operates in a low temperature environment close to absolute zero (10 mK order of magnitude, 10 mK = -273.14 °C, generally provided by the refrigerator), while its microwave control module works at room temperature. The number of control modules, in turn, increases linearly with the number of qubits.

Yan Chengyu said that the above architecture is suitable for small-scale quantum computers, and there are four drawbacks to applying to large-scale quantum computers: first, a large number of microwave control modules require extremely high costs; second, the limited physical space of the refrigerator cannot accommodate a large number of microwave modulation components (such as attenuators); third, the limited cooling power of the refrigerator cannot match the dissipated power of a large number of microwave components, which may destroy the low temperature environment required by superconducting quantum computers, resulting in a significant decline in its performance. Fourth, the distortion and delay of the waveform will occur after the microwave signal passes through a long propagation path, which especially restricts the highly sensitive quantum operation of the waveform and timing represented by the qubit gate.

"In order to solve the above four problems, it is necessary to integrate the qubits and the control module under a single chip and operate at the same temperature (generally in the order of 10mK). In this way, we can avoid using many unnecessary microwave modulation components and reduce the line connection, thus solving the first three drawbacks. Yan Chengyu introduced.

"In addition, in this integrated architecture, the drive signal is transmitted on a micron-sized chip, rather than a series of microwave components that have (negative) effects on waveform and delay, as in the traditional architecture, which has a series of microwave components that have a (negative) impact on waveform and delay, so the fourth point problem is solved." This paves the way for the development of large-scale quantum computers at the systems level. ”

According to reports, the most core part of the control module is the microwave source on the chip. Previously, several teams tried to develop an on-chip microwave source [Y.-Y. Liu et al. Science 347, 285 (2015); .C Cassidy et al., Science 355, 939 (2017)]. However, the output power of these microwave sources is in the order of 0.2 pW, which is not enough to drive a single high-quality superconducting qubit. Although they wanted to achieve greater output power, the ultra-low operating temperature of 10mk limited the actual design.

In this work, the new microwave source is an on-chip device that can be integrated with a quantum processor and is less than 1 millimeter in size. Through suitable material selection and parameter optimization, the researchers have realized a microwave source that can provide 25pW of output power, which is enough to drive 10-1000 high-quality superconducting qubits for fast quantum gate operation.

"Our device produces 100 times more power than before, enough to control qubits and perform quantum logic operations." M. Tt Nen, corresponding author of the paper and a professor at Aalto University, said.

"It's a qualitative change." Yan Chengyu said, "The number of operable qubits can be further increased by multi-cascading. In addition to the output power of the microwave source, the researchers are also concerned about the line width and noise of the output signal in the frequency spectrum: the narrower the line width, the better, and if it is too wide, it will cause crosstalk; the noise of the microwave source will cause distortion in the qubits driven by it when performing quantum computations, so the smaller the better."

"Compared to previous work, our microwave source has a narrower output signal linewidth and two orders of magnitude less noise. Based on the above results, we found that ideal superconducting qubits driven by our on-chip microwave source perform typical quantum gate manipulation with an operational distortion rate of less than 0.1% over an evolutionary time of up to 10ms. So, this on-chip microwave source does meet the needs of superconducting quantum computing. Yan Chengyu said.

In addition to the experiments, the researchers also proposed a theoretical model that can quantitatively describe the performance of its microwave source and indicate ways to further improve the performance of the microwave source, which can be used as a design guide for subsequent work.

However, the microwave source produced by the device, the continuous wave microwave source, cannot control the qubits as is. It needs to be tuned to pulse, and currently, the team is developing ways to quickly switch microwave sources on and off. Even when no pulse is formed, high-efficiency, low-noise, low-temperature microwave sources can be used for quantum technologies, such as quantum sensors.

"In addition to quantum computers and sensors, on-chip microwave sources can also serve as clocks for other electronic devices." M tt nen said, "It allows different devices to maintain the same rhythm, allowing them to operate on several different qubits in an instant of the time they need." ”

Original link: https://www.nature.com/articles/s41928-021-00680-z