#Origin Wukong Quantum Computer

[govindhtech]

China’s Origin Wukong Quantum Computer for AI Model Training

Chinese Researchers Use Quantum Computers to Fine-Tune Billion-Parameter AI Model

The state-affiliated Global Times reports that Chinese researchers optimised a billion-parameter artificial intelligence (AI) model using a quantum computer, a global first and a step towards combining quantum computing with complicated AI tasks.

Origin The experiment employed Wukong, a superconducting quantum computer with a 72-qubit chip made in the nation. Global Times said that quantum hardware fine-tuning showed that model training efficiency might be improved even with fewer model parameters, potentially alleviating computer capacity shortages.

Engineering Research Centre for Quantum Computing, Anhui

The Origin Wukong-operating Anhui Quantum Computing Engineering Research Centre reported the development Monday. The centre claims that a single batch of input data may start hundreds of quantum processes, enabling large-scale parallel computing.

Researchers refined a billion-parameter model in one trial using Origin Wukong, the magazine reported. Using psychological therapy conversations, the model reduced training loss by 15%, a key indication of learning performance. In another mathematical reasoning task, the model's accuracy rose from 68% to 82%.

According to China's People's Daily-affiliated Global Times, researchers found an 8.4% boost in training efficacy after reducing model size by 76%. Scientists believe this might lead to more energy- and memory-efficient models.

AI development often involves fine-tuning large language models (LLMs) trained for customer service or legal assistance. Adjusting millions or billions of model parameters' underlying numerical values for forecasts requires a lot of computing power. Quantum technologies may minimise traditional computer burden.

An Important Advance

Chen Zhaojun, a deputy researcher at the Hefei Comprehensive National Science Centre Institute of Artificial Intelligence, told Global Times that quantum computing has advanced. He told the magazine that this was the first time a quantum computer was used to fine-tune a large model, proving the technology can do so.

Combining quantum tools with classical systems proved effective, according to Dou Menghan, deputy director of the Anhui Quantum Computing Engineering Research Centre and vice president of Origin Quantum Computing Technology Co. “It’s like adding a ‘quantum engine,’ to a classical model,’ allowing the two to work together,” Dou told Global Times.

Over several quantum computing operations, the researchers used Origin Wukong to fine-tune. Each job maintained parallel model training, which quantum systems are ideal for. Qubits, quantum information units, allow quantum computers to do computations that classical systems cannot.

Origin Wukong's 72-qubit processor is one of China's most powerful and worldwide's most advanced, according to Global Times. Wukong has completed approximately 350,000 biomedical, financial, and fluid dynamics computations since its January 6, 2024 debut. Users in 139 countries and regions have sought remote access, Global Times said.

Wukong derives from Chinese mythology's Monkey King, Sun Wukong. Since he can transform into 72 shapes, the figure was chosen to represent the machine's variety and breadth across sectors and issue kinds.

Global Times said that industry experts believe the study may solve AI's “computing power anxiety”. Larger LLMs require more hardware, increasing costs and energy use. Quantum processors may help meet such demands even in their early stages.

Despite China's quantum research and hardware investments, global competitiveness remains heated. US, European, and Canadian quantum-AI programs have begun. Researchers behind this latest testing say Origin Wukong is one of the first quantum systems to manage a full-scale AI fine-tuning workload with demonstrable results.

Remember that the project is a demonstration, not a business usage. The story does not specify whether the researchers would publish a study on the experiment to further examine it.

[xtruss]

How Russia and China Dominate Global Quantum Computer Race

CCO//

A 50-qubit quantum computer prototype based on rubidium neutral atoms has been developed by researchers at the Moscow State University and the Russian Quantum Center.

What makes it so special?

How It Works

The new quantum computer is based on cold atoms, which is a method that involved trapping individual atoms to be manipulated with laser beams. This method allows for precise manipulation of quantum states, with the atoms being cooled and controlled via lasers.

Why Does It Matter?

Unlike Western quantum computers based on superconductors, the Russian device possesses higher stability, has lower error rates and is more energy efficient.

This technological breakthrough marks a significant leap from the 20-qubit quantum computer Russia unveiled in early 2024 and the 16-qubit quantum computer that was presented to President Vladimir Putin in 2023.

What Has China Been Doing?

Jiuzhiang, a 76-qubit quantum computer developed by scientists at the University of Science and Technology of China, became the first photonic quantum computer to achieve quantum supremacy.

The Zuchongzhi 3.0 quantum computer, also developed in China, features 105 qubits, while its predecessor Zuchingzhi 2.0 had only 66 qubits.

Another Chinese quantum computer, the 72-qubit Origin Wukong, recently reached an important milestone as it received over 20 million remote visits from more than 100 countries and regions.

How Do Western Quantum Computers Compare?

Western powers have their own achievements in the quantum computing field, which include the superconductor-based 53-qubit Sycamore quantum processor made by Google, the 27-qubit Quantum System One developed by IBM Research (first commercial quantum computer), and the 1,121-qubit Condor quantum processor also made by IBM.

Russia and China, however, are moving to outclass the Western quantum computers by developing more sophisticated and efficient systems that are less prone to making errors.

— 16th February, 2025

[nawapon17]

US Tops User List for China's Quantum Computer, Origin Wukong

[govindhtech]

Tianji 4.0: 500+ Qubit Control System from Origin Quantum

In Hefei, Anhui Province, East China, Origin Quantum Computing Technology Co. launched Origin Tianji 4.0, a fourth-generation self-developed quantum control system, furthering China's quantum technology aspirations. This new technology, announced on Tuesday, May 6, 2025, may enable more than 500 qubits, helping China's quantum computing sector industrialise and scale 100-qubit machines.

Quantum Origin Tianji4.0

Quantum computers have a quantum control system as its “neural center”. Managing qubits on a quantum device requires accurate signal production, acquisition, and control. System calibration, coherence management, and signal timing depend on these systems, which constitute the main interface between classical and quantum components.

Origin Tianji 4.0 runs Origin Wukong, China's third-generation superconducting quantum computer, better than version 3.0. The latest version improves automation, scalability, integration, and dependability.

The development team lead by Kong Weicheng, deputy director of the Anhui Quantum Computing Engineering Research Centre, enhanced quantum chip management and reading using a self-developed hardware and software architecture. This planned design reduces manual system calibration and speeds deployment cycles. Origin Tianji 4.0 and the company's four major software platforms improve superconducting quantum chip modification and testing precision and efficiency.

One of the biggest changes brought about by Tianji 4.0 is the shift from PhD-level occupations to standardised procedures that engineers can do. Shifting from rigorous hardware tweaking and expert-driven calibration to repeatable, standardised methods is crucial to preparing the system for 100-qubit quantum devices. Origin Quantum chief scientist and head of the Anhui Quantum Computing Engineering Research Centre said the launch shifts from prototype-level development to repeatable engineering production.

The framework for mass-producing quantum systems requires this capacity. Mass manufacturing is needed to build devices with higher qubit counts and operational stability for real-world applications in computationally demanding sectors.

The Origin Wukong superconducting quantum computer, powered by Origin Tianji 3.0, was active and widespread, thereby justifying the new system. Since becoming live on January 6, 2024, Origin Wukong has completed over 380,000 quantum computing operations. More than 26 million people from 139 countries and regions use it, which supports fluid dynamics, biology, and finance. The Origin Wukong platform received over 20 million visitors in March 2025, with the US, Russia, Japan, and Canada being the most active.

Tianji 4.0 is also tied to China's quantum computing objectives. The system supports the nation's next-generation quantum computing research. Origin Quantum presents its self-developed, full-stack components control systems, software, and chip integration as a “China solution” to help China overcome the engineering and technological challenges of scalable quantum computing and dominate the global quantum technology competition.

Even so little details regarding the next-generation system are available, Tianji 4.0's characteristics imply a continuous strategic focus on hardware-software co-design, system stability even with growing qubit counts, and industrial-scale deployment. The Chinese quantum ecosystem appears to be prioritising modular and high-throughput quantum systems.

The Effects

Origin Quantum's Tianji 4.0 deployment will be judged not just by technical performance measures but also by its ability to offer a repeatable basis for the larger distribution of quantum technology in China and maybe worldwide.

Quantum computing breakthroughs like Tianji 4.0 may effect adjacent sectors like artificial intelligence. Quantum computing might reduce AI training time from weeks to hours, say experts. China's first quantum computing and data medicine institution in Hefei and its successful optimisation of a billion-parameter AI model on the Origin Wukong system demonstrate the interconnection of these technical advances.

[nawapon17]

Reports: Origin Quantum Computing Unveils 72-Qubit Quantum Computer