On March 29th, Ningchang revealed its strategic roadmap for 2024, unveiling a series of strategic new products and system solutions, including the "AI Computing Stack," aimed at effectively addressing the computational support challenges throughout the entire lifecycle of large-scale model deployment.
"As large-scale model development continues to deepen, the challenges of efficiency, compatibility, and application in various industries have become increasingly complex, demanding comprehensive and full-cycle support from systematic AI computing solutions," stated Qin Xiaoning, President of Ningchang, during the press conference.
With precise anticipation of industry trends and deep insights, Ningchang introduced the "Global Intelligent Computing" strategy, aiming to provide developmental impetus to a wide range of industries through full-stack intelligent computing capabilities. This includes a comprehensive range of hardware, software, algorithms, liquid cooling, and service quality, among other dimensions of systematic AI computing solutions.
Qin Xiaoning elaborated on the "Global Intelligent Computing" strategy, stating, "Ningchang's 'Global Intelligent Computing' embodies six 'all-encompassing' characteristics, covering the entire system of hardware and software, as well as all liquid cooling products, providing full-process services from consultation to operation and maintenance, meeting the full-scenario needs of users across all industries for large-scale model development, adaptation, and deployment, and customizing professional and cost-effective AI computing solutions according to the development stage of users."
Under the guidance of this new strategy, Ningchang introduced a series of strategic new products and system solutions. Among them, the "AI Computing Stack" serves as a concentrated embodiment of Ningchang's AI full-stack capabilities, offering systematic solutions from underlying hardware to top-level application platforms, constructing a full-stack AI environment to meet the computational, storage, networking, construction, management, application, and liquid cooling requirements for large-scale model deployment.
Ningchang's CTO, Zhao Lei, explained that the global intelligent computing capability is reflected in the AI Computing Stack, showcasing both hardware strength and software optimization capabilities. On the hardware layer, various delivery forms can be flexibly combined, such as the Ningchang B8000 liquid-cooled cabinet server, which adopts a design with blind plugging for electricity, liquid, and network, reducing deployment cycles by 30 times compared to traditional methods. On the software layer, based on system engineering and algorithm models, with full-stack optimization capabilities for AI operators, services such as parallel acceleration, performance analysis, and model development optimization are provided to support AI businesses, establishing a closed-loop service for large-scale model computational power from customized computing resource allocation to model adaptation optimization, and then to efficient deployment and implementation, helping users rapidly advance AI application development and management.
To accelerate user empowerment, the first AI Computing Stack in China—Ningchang NEX AI Lab (Nettrix AI Open Lab)—has been successfully established and is open for trial in Tongxiang, Zhejiang, showcasing optimized application scenarios and industry-specific customized solutions, and providing an advanced experience of solution workflows such as large-scale models, data science, and recommendation systems through free provision of software and hardware services.
The inaugural AI computing power stack in China—Ningchang NEX AI Lab—is now officially open for business. Zhao Lei, during the introduction, stated that in the future, the "AI computing power stack" solution will further penetrate various sectors including automotive, internet, manufacturing, finance, energy, and scientific research, becoming the core engine driving digital transformation and achieving intelligent upgrades.
