Wenyi Peng

In the relentless pursuit of quantum computational advantage, we present a significant advancement with the development of Zuchongzhi 3.0. This superconducting quantum computer prototype, comprising 105 qubits, achieves high operational fidelities, with single-qubit gates, two-qubit gates, and readout fidelity at 99.90%, 99.62%, and 99.13%, respectively. Our experiments with an 83-qubit, 32-cycle random circuit sampling on the Zuchongzhi 3.0 highlight its superior performance, achieving 1×10^{6} samples in just a few hundred seconds. This task is estimated to be infeasible on the most powerful classical supercomputers, Frontier, which would require approximately 5.9×10^{9} yr to replicate the task. This leap in processing power places the classical simulation cost 6 orders of magnitude beyond Google's SYC-67 and SYC-70 experiments [Morvan et al., Nature 634, 328 (2024)10.1038/s41586-024-07998-6], firmly establishing a new benchmark in quantum computational advantage. Our work not only advances the frontiers of quantum computing but also lays the groundwork for a new era where quantum processors play an essential role in tackling sophisticated real-world challenges.

Quantum error correction (QEC) enables practical quantum computing by encoding logical qubits in many physical qubits, which can exponentially suppress the logical error rate with increasing code size provided that the physical error rate is below a critical threshold. However, the leakage of quantum information from the computational subspace presents a critical challenge to the development of scalable QEC, which creates long-lived, correlated errors that spread across space and time. Here, we demonstrate a quantum memory operating below the threshold by implementing an all-microwave leakage suppression architecture on a distance-7 surface code. We achieve a logical error suppression factor of Λ=1.40(6), definitively reversing the above-threshold scaling (Λ<1) caused by unmitigated leakage. This scheme integrates a hardware-efficient leakage reduction unit for data qubits with a fast, unconditional reset for ancilla qubits, suppressing the average leakage population after 40 cycles by a factor of 72 to 6.4(5)×10^{-4}. Our results demonstrate the viability of all-microwave control architectures for suppressing critical errors at scale, paving the way for more advanced quantum error correction implementations.

Abstract Optical neural networks (ONNs) have shown great promise in overcoming the speed and efficiency bottlenecks of artificial neural networks. However, the absence of high‐speed, energy‐efficient nonlinear activators significantly impedes the advancement of ONNs and their extension to ultrafast application scenarios like real‐time intelligent signal processing. In this work, a novel silicon/graphene ultrafast all‐optical nonlinear activator, leveraging the hybrid integration of silicon slot waveguides, plasmonic slot waveguides, and monolayer graphene is demonstrated. Exploiting the exceptional picosecond‐scale photogenerated carrier relaxation time of graphene, the response time of the activator is markedly reduced to ≈93.6 ps, establishing all‐optical activator as the fastest known in silicon photonics to knowledge. Moreover, the all‐optical nonlinear activator holds a low threshold power of 5.49 mW and a corresponding power consumption per activation of 0.51 pJ. Its feasibility and capability for use in ONNs, manifesting performance comparable with commonly used activation functions are experimentally confirmed. This breakthrough in speed and energy efficiency of all‐optical nonlinear activators opens the door to significant improvements in the performance and applicability of ONNs.

Graphene/silicon hybrid photodetector operating at communication wavelength has attracted enormous attention recently due to its potential to realize bandwidth larger than 100 GHz. However, the responsivity is intrinsically limited by the low absorption from the atomic-thick graphene monolayer, which imposes significant obstacles towards its practical application. Although plasmonic structures has been widely applied to enhance the responsivity, it may induce the metallic absorption thus limit the responsivity lower than 0.6 A/W. Twisted bilayer graphene (TBG) has been reported to hold the ability to dramatically enhance the optical absorption due to the unique twist-angle-dependent van Hove singularities. In this article, we present a design of a silicon/TBG hybrid photodetector with a responsivity higher than 1 A/W and bandwidth exceeding 100 GHz. The enhanced responsivity is achieved by tuning the twisted angle of TBG to increase the absorption within the 1550 nm as well as utilizing the silicon slot waveguide to boost the mode overlap with TBG. The fabrication process of proposed design is also discussed demonstrating the advantages of low fabrication complexity. The proposed silicon/TBG photodetector could not only exhibit superior performance compared to previously reported silicon/monolayer graphene photodetector, but also pave the way for the practical application of graphene-based silicon optoelectronic devices.

We proposed an automatic polarization controller based on a novel thermal phase tuning structure on the silicon platform, with a polarization control speed of up to 20 krad/s, which is the fastest reported silicon-based device.