Solving quantum computing problems – Nature Electronics

Quantum computers can be used to solve complex problems that are impossible for the most conventional computer. But building a working quantum computer is, of course, not easy. To start, you need a scalable qubit platform. You need to be able to control and calculate the state of the qubits. You need to be able to connect multiple qubits. And you need to know if mistakes have occurred and correct them. Each of these jobs has its challenges. However, researchers in academia and industry continue to find solutions, which bring robotics closer to reality.¹. This matter of Natural Electronicswe highlight some of the latest developments.

Superconducting qubits are currently one of the leading platforms in the race to build a functional quantum computer. However, such a machine would require a million or more physical qubits^2,3and each qubit uses an individual signal line to control it. This creates a challenge – more qubits means more wires. Platforms often have limited integration as qubits operate at millikelvin temperatures and room temperature control electronics. In an article appearing in this issue, Shu-Jen Han and colleagues report a quantum processor unit in which superconducting qubits and single-flux quantum control electronics are integrated into a single multi-chip module via flip-chip bonding. In this approach, the qubit layer and control electronics can operate in the same millikelvin temperature range. Researchers – based at Seeqc Inc. New York and Seeqc UK in London – they also use digital demultiplexing to distribute control pulses to multiple qubits and, with this, break the same amount of control lines into the number of qubits.

Distributed quantum networks – where many chips are connected via communication channels – could help develop quantum computing. Even microwave technology, which is the basis of modern communications, can help create such a network. But microwave photons are sensitive to thermal noise, which often destroys quantum state information during transit. In another article this month, Jingjing Niu, Youpeng Zhong, Dapeng Yu and colleagues – based at the International Quantum Academy in Shenzhen, Southern University of Science and Technology in Shenzhen, Ningxia University, and Hefei National Laboratory – report a thermal-noise stable microwave quantum network. This method removes the operating temperature of a millikelvin qubit from the communication channel, which operates at 4 K. As Peter Rabl of the Technical University of Munich notes in an article accompanying News & Views, “The protocol is also not limited to 4 K, and can be combined with high-temperature superconductors that operate at a water-nitrogen temperature of 77 K.”

Silicon spin qubits can exploit the manufacturing techniques used to build conventional computers and thus provide a promising route to scalable quantum processors. Calculations of such qubits usually involve radiofrequency single-electron transistors. But these have a trade-off between chip space and reliability. In another article in this issue, Jacob Chittock-Wood, John Morton, Fernando Gonzalez-Zalba and colleagues report a radiofrequency electron cascade readout method for coupled spin qubits. The team – based at the University of London, Quantum Motion, the University of Cambridge, the University of Oxford and imec – uses a method to count two electron spins in a silicon planar metal-oxide-semiconductor quantum dot array.

A large-scale network of spin qubits will also require methods to identify quantum components at scale. To address this, Giordano Scappucci and colleagues at Delft University of Technology report a crossbar chip for benchmarking semiconductor spin qubits. Another challenge in scaling up semiconductor spin qubits is the delicate tuning required to achieve and maintain qubit performance. To address this, Jonas Schuff, Natalia Ares and colleagues at the University of Oxford, the University of Basel and the Mind Foundry in Oxford report an independent tuning process for spin qubits.

Effective quantum communication probably requires qubit error rates of less than 10^-10 (ref. ⁴). However, errors are inevitable in physical qubits, and the error rates of spin qubits fall short of this limit. Therefore, it is important to recognize errors and correct them. In another article this month, Guangchong Wang, Guangchong Hu, Yu He, Dapeng Yu and their colleagues report the discovery of quantum defects in a silicon quantum processor. The researchers – based at the Southern University of Science and Technology in Shenzhen, the International Quantum Academy in Shenzhen, and the Hefei National Laboratory – use a system made of four spin qubits and one electron spin qubit, and show that a single qubit error can be detected. (See also the accompanying article News and Opinions about the work from Lieven Vandersypen of Delft University of Technology.)