Interview: Mr. Nathan Lacroix and Dr. Sebastian Krinner
Can you tell us about your current research activities?
Sebastian: We develop the basic building blocks of future quantum computers based on superconducting circuits. As in any quantum system, the quantum states in these devices are fragile, and quantum computers will therefore need to implement error correction. At the moment we develop and improve important ingredients for quantum error correction, ranging from high-fidelity readout and gate operations to crosstalk mitigation and parallelization of operations. At the same time, we bring together all building blocks early enough to find out what presently limits the performance of the error correction cycles.
What is especially exciting about this topic in your view?
Sebastian: This research allows us to answer the question of whether fault-tolerant quantum computers are practically realizable at all. According to the theory of quantum fault tolerance developed in the nineties, a quantum computer can work in principle given some assumptions on the noise of the device. Now we have the opportunity to test this with actual devices that come with real-world noise. It's part of the fascination with this research area that we can put to the test important theoretical concepts.
Nathan: Quantum computers are expected to solve problems that classical computers cannot solve in practice, but we are relatively confident that many of these tasks will require error correction. Therefore, developing error correction is essential to unlocking the full potential of quantum computers. With superconducting qubits, we have a platform that allows us to test all different elements of error correction and combine them in ways that have not been demonstrated before. There is still a long way to go to achieve fault-tolerant quantum computing, but being part of this journey is exciting.
What difficulties did you need to overcome to achieve your recent results on a 17-qubit quantum processor?
Nathan: It was a large team effort to bring together chip design, fabrication, cryogenics, room-temperature electronics, testing and characterization of the chip, and finally to run the experiment – up to ten people worked on the experiment at the same time. A major, specific technical challenge was to improve our two-qubit gate fidelity by carefully compensating the distortions of signals applied to the qubits for their control. This allowed us to suppress the effects of imperfections in the analog signal path.
Sebastian: It was also an integration effort. We had to bring together all previously developed components and improve them at the same time. Parallelizing operations was a particular challenge, starting with tune-up procedures and extending that to two-qubit gates.
What do you view as the next challenge when scaling up the number of qubits?
Sebastian: In terms of device fabrication, it will be necessary to master 3D integration to make it possible to guide signals to the 2D qubit lattice from the third dimension. We also want to improve the qubit quality in terms of coherence times and number of spurious modes. Developing multiplexing techniques could be important to control the quantum processor. From a design standpoint, we see some potential for optimizing circuit parameters further to achieve a shorter quantum error correction cycle.
How does the Zurich Instruments QCCS support your research?
Nathan: An additional challenge we face when building quantum computers is the scaling of the room-temperature electronics. These control electronics need to become more compact if they are to control 100 qubits and beyond. Previously, we developed most of the electronics by ourselves. Since 2015, a growing, close collaboration with Zurich Instruments has allowed us to focus on other aspects of our research while providing feedback on the features we consider important for the next generation of instruments. We are then the first ones to test these new features.
Sebastian: We use Zurich Instruments' equipment for low-noise and high-resolution control signal generation as well as for readout with FPGA-based fast signal processing. If we had to build control electronics on a large scale with similar characteristics in terms of noise level, synchronization and phase stability, that would distract us from our core research activities.