Research On Low-latency Readout And Real-time Feedback Control System In Superconducting Quantum Computing

The performance of quantum computers far exceeds that of classic computers on some specific problems.For example,the SHOR algorithm designed for large number decomposition and the GROVER algorithm designed for data search show amazing acceleration capabilities compared to classic algorithms.Quantum computers provide a new way to solve a specific problem,and their computing power is not limited to Landauer’s Principle,which has great potential for applications.Hence,the development of quantum computers has become an important battlefield of scientific and technological competition on the international stage.With sufficient qubit fidelity,the performance of a quantum computer increases exponentially with the number of qubits.However,in the current noisy intermediatescale quantum-(NISQ)stage,the qubit fidelity is limited by noise due to the intrinsic decoherence nature of physical qubits.Going beyond the NISQ era requires the use of quantum error correction circuits(QECs)to cancel noise in real time.QECs utilize multiple physical qubits(functionally divided into data qubits and ancilla qubits)to encode a logical qubit,which corrects errors in data qubits in real time based on the measurement results of ancilla qubits during the coherence time of qubits.QECs require electronics to provide excellent scalability and support dynamic quantum circuits.To take a step toward to QECs,this paper studies the readout and feedback electronics in QECs,and the main contents include the following:1.Developed measurement and control management software.Because the firstgeneration electronics lack corresponding measurement and control management software,this study simplified the system expansion,improved the control accuracy,and ensured the experimental efficiency by means of instrument virtualization,channel calibration,and communication optimization.Developed measurement and control management software for first-generation electronics.2.Proposed a distributed hardware architecture.Aiming at the problem of insufficient scalability of the existing solutions,this paper proposes a distributed hardware architecture.The system based on this architecture takes the chassis as the expansion unit,which not only improves the integration of the system but also reduces the complexity of wiring while achieving high scalability.The processing node of the architecture is implemented by FPGA,which can simultaneously integrate readout and feedback electronics,providing great flexibility for scale and function expansion.3.Designed and implemented low-latency readout electronics.To realize lowlatency closed-loop feedback in dynamic quantum circuits and meet the requirements of dynamic quantum circuits for readout latency performance and dynamic readout function,this paper designed and developed low-latency readout electronics.Based on the idea of multi-channel parallelism and single-instructionmultiple-data,a low-latency readout algorithm is designed,and a codeword trigger mechanism is introduced for readout electronics,thus realizing the support for dynamic quantum circuits.In addition,a matched filter is used to maximize the input signal-to-noise ratio,and 2 GB of data storage is provided to alleviate the efficiency issues caused by frequent communication.4.Designed and implemented real-time feedback electronics.To meet the requirements of the arbitrary workflow of circuits in dynamic quantum computing,this study extends the timing control instructions on the basis of the RISC-V instruction set and implements a timing control unit(TCU)for dynamic quantum circuits.By programming the TCU,the dynamic quantum circuits can be controlled in runtime with deterministic timing.In addition,low-latency data links are developed to provide communication solutions for system expansion.Research on management software was applied to the quantum computing system of China’s first online cloud platform in 2018.Research on the new generation of readout and feedback electronics has been successfully applied to active reset circuits based on the repeat-until-success(RUS)strategy.The study verified the support of the new generation of measurement and control electronics for dynamic quantum circuits through the T1 measurement experiment of two qubits.The innovations of this paper are as follows:1.Proposed a distributed hardware architecture.Hardware adopts a modular design idea,and the functional integration of different electronics can be realized by combining different hardware.The readout and control module is installed in a chassis,and by configuring the electronic modules in the chassis,the system can adapt to different quantum computing scenarios.The chassis can be cascaded through low-latency data links,allowing the system to support scaling in excess of hundreds of qubits.2.Designed and implemented low-latency readout electronics.The readout electronics achieve a maximum sampling rate of 2 Gsps,an input bandwidth of 600 MHz,and an electronics readout delay as low as～40 ns.In each experiment,it can support more than 10 million readout operations and up to 16 frequency points in each readout operation.The readout electronics lay the foundation for dynamic quantum computing with sufficient performance.3.Designed and implemented a TCU for controlling dynamic quantum circuits.The TCU contains classical instructions for arithmetic,loops,and jumps,as well as timing control instructions for qubit manipulation.The TCU can not only functionally support dynamic quantum circuits but also improve the efficiency of experiments in terms of performance.In the T1 measurement experiment based on the RUS strategy,the dynamic circuit based on the TCU achieves more than 5 times the efficiency improvement compared to the static circuit.