Breakthrough in Quantum Computing: Jiuzhang 4.0 Surges Ahead

By VirentaNews Staff — May 16, 2026

China’s Jiuzhang 4.0 photonic quantum processor has achieved a new milestone in quantum computing by successfully implementing Gaussian boson sampling with 1,024 squeezed states across 8,176 optical modes. This breakthrough demonstrates a level of computational complexity that exceeds the capacity of even the most powerful classical supercomputers, reinforcing the potential for photonic systems to achieve practical quantum advantage. The experiment, published in Nature, marks a significant advance in scalability, fidelity, and programmability of quantum photonic hardware, setting a new benchmark for quantum supremacy demonstrations.

Unprecedented Scale of Quantum Optical Circuitry

The core achievement of Jiuzhang 4.0 lies in its ability to generate and manipulate 1,024 high-efficiency squeezed states within a hybrid spatial–temporal encoded circuit comprising 8,176 optical modes. This represents a fivefold increase in photon number and a threefold expansion in mode count compared to its predecessor, Jiuzhang 3.0. Each squeezed state was produced with over 95% efficiency using ultra-low-loss photonic integrated circuits and time-multiplexed detection, enabling precise control and measurement. The system performed Gaussian boson sampling (GBS) — a task proven to be intractable for classical algorithms — in approximately 200 seconds. In contrast, researchers estimate that Frontier, the world’s most powerful classical supercomputer, would require over 100 million years to simulate the same output distribution. These figures, derived from cross-validated statistical benchmarks, underscore a quantum computational advantage exceeding 10¹⁸ in runtime efficiency.

Key Players Behind the Quantum Leap

The Jiuzhang 4.0 project is led by a team at the University of Science and Technology of China (USTC) in Hefei, under the guidance of quantum physicist Jian-Wei Pan, a pioneer in quantum communication and photonic quantum computing. Collaborators include researchers from the Shanghai Institute of Microsystem and Information Technology and Tsinghua University, who contributed to the design of superconducting nanowire single-photon detectors and low-loss photonic chips. The team leveraged advances in silicon photonics and machine-learning-optimized interferometer calibration to stabilize the massive optical network. Unlike competing quantum platforms such as superconducting qubits or trapped ions, the USTC group has focused exclusively on photonic architectures, betting on their inherent stability at room temperature and compatibility with existing telecom infrastructure. Their sustained investment in GBS as a pathway to quantum advantage has now yielded the most convincing evidence to date that non-universal quantum devices can outperform classical counterparts in well-defined computational tasks.

Trade-offs in Photonic Quantum Advantage

While Jiuzhang 4.0 demonstrates extraordinary performance in Gaussian boson sampling, its computational utility remains specialized. Unlike gate-based quantum computers aiming for universal computation, photonic GBS machines excel only in sampling problems with applications in graph theory, molecular vibronic spectra simulation, and machine learning kernel methods. The system’s lack of dynamic programmability beyond interferometer reconfiguration limits its adaptability. Moreover, scaling further will require breakthroughs in photon loss mitigation, as even minor losses compound exponentially in large circuits. On the other hand, the use of room-temperature optics avoids the costly cryogenic systems needed for superconducting qubits, potentially lowering operational barriers. The photonic approach also offers natural resilience to decoherence and seamless integration with fiber-optic networks, positioning it as a viable platform for future quantum internet applications. However, translating GBS advantage into practical industrial use cases remains a significant challenge.

Why This Breakthrough Comes Now

The success of Jiuzhang 4.0 is the result of over a decade of incremental progress in quantum photonics, but recent advances in integrated circuit fabrication and single-photon detection have been decisive. Improvements in lithographic precision have enabled the creation of kilometer-scale waveguide circuits on silicon chips with propagation losses below 0.1 dB/m, a critical threshold for maintaining quantum coherence across thousands of modes. Concurrently, machine learning techniques have been deployed to calibrate and stabilize the phase settings of the interferometric network in real time, compensating for thermal drift and manufacturing imperfections. These developments, combined with the maturation of superconducting nanowire detectors with near-unity efficiency, have collectively overcome previous bottlenecks in scalability and measurement fidelity. The timing also reflects intensified global competition in quantum technologies, with China, the U.S., and the EU pouring billions into quantum research, accelerating the pace of innovation.

Where We Go From Here

In the next 6 to 12 months, three scenarios are likely. First, independent verification of Jiuzhang 4.0’s results by research groups in the U.S. or Europe may confirm the quantum advantage claim, potentially leading to broader acceptance of photonic platforms in quantum computing roadmaps. Second, efforts will intensify to find practical applications for GBS, particularly in quantum-enhanced machine learning and materials simulation, possibly yielding niche industrial deployments. Third, rival teams may attempt to simulate the experiment using optimized tensor network algorithms or exascale computing, testing the boundaries of classical tractability. Regardless of the outcome, the bar for quantum supremacy has been raised significantly, forcing a reassessment of what constitutes meaningful quantum advantage beyond synthetic benchmarks.

Bottom line — Jiuzhang 4.0’s demonstration of Gaussian boson sampling at an unprecedented scale represents a landmark in quantum computing, proving that photonic systems can achieve computational dominance while highlighting the ongoing challenge of translating theoretical advantage into real-world utility.

Source: Nature