Jiuzhang 4.0 and the Race for Photonic Quantum Advantage

The system is designed to run a specialized problem called Gaussian boson sampling (GBS)—a probabilistic task involving many photons passing through a complex interferometer. The output distribution is extremely difficult for classical computers to simulate as the number of photons increases.

In the Jiuzhang 4.0 experiment, researchers:

• Injected 1,024 high‑efficiency squeezed states of light into the processor
• Used a programmable photonic circuit with 8,176 optical modes
• Detected up to 3,050 photons in sampling events

These design choices allow the machine to generate extremely complex optical interference patterns that encode the sampling problem.

The new quantum computing record

Jiuzhang 4.0 set what researchers describe as a world record for optical quantum information experiments by dramatically increasing the number of photons involved in a Gaussian boson sampling test.

According to the reported results, the prototype solved the sampling task more than 10^54 times faster than the most powerful classical supercomputers under certain modeling assumptions.

While such comparisons depend on the exact algorithms and simulation methods used, the experiment is presented as evidence of robust quantum computational advantage—meaning the quantum system performs a task that classical simulations struggle to reproduce even as algorithms improve.

How it advances beyond Jiuzhang 3.0

Jiuzhang 4.0 is the latest system in a series of photonic quantum prototypes from the same research team.

The previous model, Jiuzhang 3.0, demonstrated 255 detected photons in a boson sampling experiment.

The new prototype expands several key dimensions:

Photon scale

• Jiuzhang 3.0: 255 photons detected
• Jiuzhang 4.0: up to 3,050 photons, more than ten times larger

System architecture

• Jiuzhang 4.0 uses 1,024 squeezed‑state sources and an 8,176‑mode programmable optical network, greatly increasing the complexity of the photonic circuit.

Programmability

• The newer processor is designed as a programmable photonic platform, enabling different configurations of optical transformations rather than a single fixed experiment.

These changes make the device capable of generating much larger quantum states and sampling distributions than earlier photonic machines.

Why Gaussian boson sampling matters

Gaussian boson sampling is not a general‑purpose computing task. Instead, it is used as a benchmark problem to test whether quantum devices can outperform classical algorithms.

The difficulty arises from the mathematics required to calculate probabilities of photon detection patterns. For large systems, these probabilities depend on computing functions such as loop hafnians, which become exponentially hard as the number of photons grows.

As a result, increasing photon count and optical modes quickly pushes classical simulation beyond practical limits. Demonstrations like Jiuzhang 4.0 therefore serve as experimental evidence that quantum systems can reach regimes where classical computers struggle.

What this milestone means for quantum computing

Jiuzhang 4.0 shows that photonic quantum hardware can scale to thousands of photons while maintaining enough fidelity to demonstrate quantum advantage in sampling tasks.

That matters for several reasons:

• Stronger evidence for quantum advantage: Larger experiments make it harder for classical algorithms to catch up.
• Progress in photonic architectures: Photons are attractive for quantum computing because they operate at room temperature and can travel long distances in optical systems.
• Foundations for future photonic quantum machines: Scaling sources, detectors, and optical circuits is essential for eventually building fault‑tolerant systems.

However, Jiuzhang 4.0 is still not a universal quantum computer. It performs a specific type of sampling experiment rather than running arbitrary algorithms.

The bigger picture

The Jiuzhang series highlights a major global strategy in quantum computing: multiple hardware approaches competing to reach scalable quantum advantage. Superconducting, trapped‑ion, neutral‑atom, and photonic systems all pursue different paths toward the same goal.

Jiuzhang 4.0’s 3,050‑photon experiment represents one of the largest photonic demonstrations to date, pushing the boundary of what optical quantum processors can achieve. Whether such systems can evolve into fault‑tolerant, general‑purpose quantum computers remains an open research challenge—but the experiment shows that photonic quantum technology is rapidly advancing toward that possibility.