AI Inclusiveness・Ecological Co-integration | The Lenovo AI Workstation Ecosystem Cooperation Salon Concluded Successfully

In 2025, artificial intelligence technology is comprehensively penetrating into key scenarios of various industries in a disruptive manner. From underlying computing power support to upper - layer application innovation, AI has become the core engine driving enterprises to reconstruct production efficiency and business models. As the wave of the popularization of large models sweeps in, the intelligent transformation of enterprises has also entered a critical window period from "technological experimentation" to "in - depth implementation" — a series of issues such as how to break the threshold of computing power deployment, balance data security and efficiency improvement, and build a sustainable ecological cooperation model have become common challenges faced by all industries.

With the goal of helping enterprises achieve intelligent transformation "painlessly", Lenovo, with the core concept of "scenario - driven, security - based, and ease - of - use first", has taken the lead in launching the industry's first AI workstation solution pre - installed with the DeepSeek large model. With a highly user - friendly local deployment method and cost, it provides an implementation model of "scenario - driven, security - based, and ease - of - use first" for the intelligent transformation of enterprises, and explores a new path that combines technological inclusiveness and ecological symbiosis for the transformation of AI technology from theoretical innovation to industrial value.

On the occasion of this technological innovation, Lenovo, together with 36Kr and NVIDIA, initiated a sub - venue event of the "36Kr AI Partner Conference" in Shanghai on April 18th, bringing together industry leaders and front - line investors to conduct in - depth discussions around "the innovation path and ecological co - construction of enterprises in the era of AI popularization".

Lu Yuan, the Marketing General Manager of Lenovo's AI Workstation; Song Yiming, the Semiconductor Architect of NVIDIA; Gu Xuguang, the Senior Director of Lenovo's "Qingtian" AI Technology Center; Ye Shuaimin, the Senior Investment Director of Lenovo Capital and Incubation Group; and Wan Yang, the General Manager of Lenovo's AI PC for Government and Enterprise Business and other industry experts, as well as enterprise representatives from various fields gathered at the event.

At this salon, the guests present conducted in - depth sharing and exchanges based on their respective businesses and professional fields.

Lu Yuan, the Marketing General Manager of Lenovo's AI Workstation introduced Lenovo's "device - edge - cloud - network - intelligence" layout based on the full - stack AI strategy: at the device side, Lenovo creates AIPC products centered around the "Tianxi Personal Intelligent Agent System"; at the cloud side, the "Wanquan Heterogeneous Intelligent Computing Platform" covers servers, storage, data networks, etc.; at the service side, the "Qingtian" Intelligent IT Engine provides industry intelligent agent solutions. Lu Yuan emphasized that Lenovo has a rich computing power product matrix, covering multi - level computing power requirements at the device side. Lenovo's AI workstation can support AI inference applications of large models from 7B to 70B, and can achieve a 2 - hour rapid deployment with NVIDIA's Ada series GPUs, allowing users to "quietly" enjoy AI computing power. Lenovo's AI workstation can meet three major needs of enterprises during the private deployment of large models: ensuring enterprise data security, quickly verifying input - output ratios, and rapidly building environments and ensuring asset security.

In terms of ecological cooperation, Lu Yuan introduced that Lenovo provides three cooperation models for ecological partners: being equipped with Lenovo's computing power base, obtaining Lenovo's marketing resource support through the co - branding model, and integrating products under the Lenovo brand into the core sales system.

Song Yiming, the Architect of NVIDIA's Professional Visualization Team shared the development history and the latest technological progress of GPUs. He revealed that the next - generation Blackwell GPU technology has key upgrades. It not only supports graphics card segmentation, but also introduces the DP2.1 interface (single - port supports 8K) for the first time on professional cards, and supports 96GB of video memory and YUV422 codec. In addition, NVIDIA has launched the DGX Spark for desktop - side inference applications, which is specifically designed for local deployment of inference models by enterprise individual users.

When talking about the cooperation with Lenovo's AI workstation, he pointed out that the most concerned factor in DeepSeek inference is the capacity of video memory. Starting from customers' needs in terms of performance, price, etc., NVIDIA's professional cards can better balance the overall performance stability, power consumption, power supply and heat dissipation.

Gu Xuguang, the Senior Director of Lenovo's "Qingtian" AI Technology Center in the Solution and Service Business Group pointed out that Lenovo's AI all - in - one machine and AI workstation are not simply hardware products, but integrated intelligent agent platforms that integrate a complete software stack and underlying solutions. When applied within enterprises, on the one hand, Lenovo can solve various possible problems, provide a good user experience, evaluation and testing, and enterprise - level functions; on the other hand, focusing on general application scenarios such as office work, operation, customer service, sales, and marketing, it can meet the diverse needs of enterprises. Regarding the most difficult after - sales problem of the all - in - one machine, Gu Xuguang emphasized that relying on Lenovo's national after - sales network, it can achieve rapid replacement of various accessories and remote operation and maintenance.

In response to the on - site audience's question about industrial data security, Gu Xuguang pointed out that Lenovo's adaptation center can provide a physically isolated test environment and data desensitization tools to support customers in offline testing and ensure industrial data security.

Ye Shuaimin, the Senior Investment Director of Lenovo Capital and Incubation Group shared the investment logic and opportunities in the era of AI popularization, pointing out that "silicon - based - code - data - intelligent agent" are the four core production materials. He believes that the generalization of AI may change the traditional customized model of enterprise software, and AI is shifting from focusing on processes to focusing on the improvement of individual productivity. There are more opportunities in the To B market in scenarios where AI can significantly improve individual efficiency. At the same time, the business model of delivering by results is expected to subvert the traditional software industry logic of "selling manpower" and customized development. Regarding entrepreneurship, he believes that one should avoid starting a business in the direct competition field of large models, and the core is to master dynamic endogenous data and build a self - contained scenario barrier.

When answering questions from the on - site audience, Ye Shuaimin said that Lenovo Capital and Incubation Group especially focuses on AI - first projects. For example, for teams developing vertical - field intelligent agents based on Lenovo's workstations, Lenovo Capital and Incubation Group is willing to provide dual support of angel - round investment and ecological resources.

In the wave of AI popularization, Lenovo will continue to strengthen the AI innovation ecosystem with its ecological partners. Relying on the technical base of AI workstations, it will jointly build a full - chain collaborative system of "computing power - model - scenario" with large - model manufacturers, hardware partners, and ISVs, co - build AIGC application scenarios, promote the large - scale implementation of intelligent agents in all industries, form a closed - loop of "technological innovation - industrial verification - value feedback", and support enterprises to achieve long - termism in the intelligent era from all dimensions.

In the future, Lenovo will deepen computing power cooperation with NVIDIA, launch an upgraded solution supporting the next - generation Blackwell architecture to reduce the deployment cost of large models; introduce 500 ecological enterprises through the "Star Project", accelerate project implementation through joint R & D and dual - brand marketing; upgrade the "Qingtian" engine to version 4.0 to achieve cross - model orchestration and automatic generation of knowledge bases...

Shouldering the responsibility of the "chain leader", Lenovo will, as always, promote the integration of large, medium, and small enterprises, build a new "two - way empowerment" ecosystem, make AI the core driving force for industrial upgrading, and embark on a new "intelligent +" future with its partners.