In the ICT business and computing field, Huawei's goal is to become "NVIDIA." Its core advantage lies not in having its own proprietary large model, but in supporting customers in building excellent large models.
News on the 16th: Huawei's Xinsheng Community has published a summary of a discussion between Chairman of the Supervisory Board Guo Ping and new employees. During the discussion, Guo Ping touched on multiple core topics, including Huawei's future strategy, artificial intelligence layout, chip competitiveness, 6G progress, and views on talent.
Looking to the future, Guo Ping said that Huawei's core strategy is "focus," that is, deepening its work in the "connectivity" and "computing" fields in which the company excels, with no plans for business expansion. Huawei hopes to achieve better performance, provide digital infrastructure solutions to global customers, offer diversified solutions for AI development, and contribute to the rise of China's electronics industry. Huawei's revenue has now recovered to its historical peak level. In 2020, Huawei achieved sales revenue of 891.4 billion yuan, setting a historical record, and in 2025 it achieved global sales revenue of 880.9 billion yuan, approaching the historical peak.
Speaking about business layout in the AI era, Guo Ping regards artificial intelligence as one of Huawei's greatest opportunities. He believes that the core players globally in computing power, talent, and data are mainly from the United States and China, and Huawei can represent China's highest level and provide a solid choice for China and the world. Huawei has business areas such as smartphones and intelligent driving that possess a data flywheel model of technological iteration, and building world-class intelligent agents with global competitiveness through artificial intelligence is an important opportunity.
Regarding proprietary large model strategies for different business segments, Guo Ping gave a differentiated layout. In the ICT business and computing field, Huawei's goal is to become "NVIDIA." Its core advantage lies not in having its own proprietary large model, but in supporting customers in building excellent large models, ensuring that any large model in the world can run efficiently on Huawei Ascend, Kunpeng, super nodes, and clusters. For the terminal business and intelligent driving business, proprietary large models are more advantageous, because most mainstream open-source large models only open-source their weights, and enterprises face challenges in subsequent additional training or fine-tuning, and there is lag in open-sourcing data. For Huawei Cloud's business, as cloud and artificial intelligence deeply integrate, the lack of a competitive large model will lead to a lack of business competitiveness, so Huawei Cloud must have its own proprietary large model to maintain competitive advantage.
Regarding chip competitiveness, Guo Ping acknowledged that Huawei is constrained in advanced processes, so it is even more necessary to promote the integration of design and manufacturing and leverage comprehensive advantages to make up for process shortcomings. He cited the "Tao's Law" proposed by He Tingbo, whose essence is replacing "geometric scaling" with "time scaling," closely combining separated process and manufacturing to enhance competitiveness under constrained conditions. Guo Ping believes that in the AI era, accelerated technological iteration may weaken moats such as NVIDIA's CUDA ecosystem. Under process constraints, the Huawei Ascend 950 series narrows the gap through architectural innovation, adopts system architecture thinking to reorganize competitive elements, and achieves overall advantages by developing clusters and super nodes.
In the carrier business, Guo Ping said this is Huawei's most globally competitive business, with the highest input-output efficiency. Facing 6G, Huawei is committed to making a difference in "from 0 to 1" exploration, while practicing "more, faster, better, and more economical" with strong execution. Huawei hopes to promote a unified global standard and continue to be an important contributor to, implementer of, and participant in market success for that standard.
Regarding the Vehicle BU business, Guo Ping said Yinwang is a company controlled by Huawei and connected through capital with many partners, and the intelligent automotive component business it provides is an extension of Huawei's ICT business. Huawei's intelligent automotive component business has been global from the outset, but complete vehicles have notably regional and geopolitically sensitive characteristics. Huawei chooses "not to build cars," and whether automakers choose to sell only in China or go global depends on each automaker's own strategic choice.
Speaking about AI's impact on talent, Guo Ping believes that in the future society, a very small number of top geniuses may devote themselves to "creating AI," while most people will be "using AI." The scarcest ability in using artificial intelligence is not mastering some programming skill, but the ability to accurately describe and ask questions. Therefore, talent in literature and philosophy is increasingly favored. He advised new employees to have the ability to keep learning, because the knowledge learned in university may already be obsolete on the day they join the company.
Regarding supply chain management, Guo Ping said that under the current constrained supply situation, supply is the bottom line of guarantee. Huawei needs to manage "timely complete-kit rate" and "procurement lead time" well, accelerate turnover, reduce inventory, and ensure that no link in the supply chain falls behind. Huawei does not take surpassing a particular company as its development goal, but focuses on doing its own work well. In the supply chain field, Huawei needs to continue learning from Apple, including the competitiveness, flexibility, and resilience of its supply chain.
During the discussion, Guo Ping also talked about Huawei's self-criticism culture. He believes that self-criticism originates from personal cognition, not a "struggle session," and is based on reviewing events that have already occurred, aiming to summarize lessons learned and clarify future directions for improvement. He emphasized "self" rather than "criticism," and the subject must be the individual actively recognizing problems. As for how new employees can maintain long-term motivation, Guo Ping suggested that joining Huawei should come from a love of the profession, because only love can support perseverance to the end.
Disclaimer: The article comes from the internet. If there is any dispute, please contact customer service.
