Xunce (03317) announced that it has recently entered into separate strategic cooperation agreements with ILUVATAR COREX, BIREN TECH, and another leading AI chip company. These strategic partnerships mark a significant step forward for the company in the integration of "computing power + data" and represent a key move to deeply merge its data infrastructure platform, data tokenization capabilities with computing power, thereby accelerating the large-scale application of enterprise-level AI.

The company has a deep presence in vertical sectors including finance, telecommunications, power, energy, advanced manufacturing, biomedicine, smart cities, embodied AI, commercial aerospace, low-altitude economy, and smart vehicles, amassing extensive experience in scenario-specific data processing and tokenization. Computing power providers like ILUVATAR COREX and BIREN TECH possess core technical advantages in chip architecture and computing power optimization.

As AI evolves rapidly from general training to vertical inference, the company's collaborations with these three parties aim to deeply couple scenario-specific data capabilities with computing power. Together, they will develop "software-hardware integrated" solutions tailored for vertical industries, driving the accelerated deployment of computing power in specific application scenarios.

Under the strategic cooperation agreement with ILUVATAR COREX, the two parties will jointly conduct chip-level adaptation and optimization for vertical scenarios, creating efficient, integrated software-hardware computing solutions. Leveraging the company's deep industry expertise and vast experience in business data governance across sectors like finance, manufacturing, and energy, combined with ILUVATAR COREX's technological leadership in general-purpose GPU and AI chip architecture design and computing performance optimization, they will carry out joint chip-level testing and deep optimization for typical industry scenarios to form integrated solutions for specific business needs.

Furthermore, they will jointly research heterogeneous computing network solutions for complex business scenarios. By building a coordinated "cloud-edge-end" computing power scheduling system, they aim to achieve unified management and dynamic allocation of diverse heterogeneous computing resources, accelerating the large-scale application of intelligent computing in areas like financial risk control, industrial quality inspection, and smart energy.

The two sides will also collaborate on application R&D for embodied AI (physical AI) in scenarios such as intelligent manufacturing, smart services, and intelligent inspection, jointly exploring commercialization paths for physical AI.

Pursuant to the cooperation memorandum with BIREN TECH, the two parties will focus on the intelligentization needs of key areas like urban management, collaboratively developing end-to-end intelligent computing solutions covering both hardware and software to break through core technical bottlenecks.

They will jointly invest in building intelligent computing clusters, integrating BIREN TECH's GPU chip computing power advantages with the company's data computing platform capabilities to establish an efficient and stable computing power support foundation.

Additionally, they will jointly develop specialized integrated machines for fields like urban management, optimizing product adaptability and deployment convenience to promote market penetration and large-scale application in key industries, achieving a "ready-to-use" deployment goal.

Cooperation will also extend to the development of vertical application intelligent agents, fine-tuning model capabilities with structured data, and the debugging, verification, and combined validation of various models.

According to the strategic cooperation agreement with the other AI chip firm, the two parties will explore full-stack intelligent computing end-to-end technical paths and solutions covering software and hardware for key vertical fields such as intelligent manufacturing and tech finance.

By combining the GPU chip computing power advantages of the partner with the company's data computing platform capabilities, they will jointly explore computing power infrastructure construction plans to build an efficient and stable computing power support base.

Simultaneously, they will conduct adaptation and verification in areas including the development of vertical application intelligent agents, fine-tuning model capabilities with structured data, and debugging various models, promoting product market expansion and large-scale application in key industries to realize the goal of "ready-to-use" deployment.