The immense innovative potential of 人工智能 positions it as a pivotal variable for growth. Currently, the competitive focus within the field is shifting from "technology research and development" towards "value realization." How is this value realized? Industry, computing power, and talent hold the keys to unlocking this growth. Chongqing has already established platforms like the "Two Institutes and One Bay" to build a highland for 人工智能 innovation and application.
The "Two Institutes and One Bay" initiative sees the "Two Institutes" focusing on talent and technological sourcing, while the "One Bay" concentrates on scenario development and industry incubation. Together, they create a complete closed-loop for the "value conversion" of 人工智能. The city's Sixth Plenary Session of the Ninth Municipal Party Committee made clear deployments for deepening and expanding "AI innovation application." How will this most futuristic piece of Chongqing's innovation landscape transform and leap forward?
Starting with the "Two Institutes and One Bay," this is the core base for Chongqing's 人工智能 industry layout. Its central mission is to provide a real proving ground for the innovation and implementation of 人工智能 technologies. At the heart of driving this process are the two institutes within the Bay: the Chongqing 人工智能 Institute and the Chongqing General 人工智能 Research Institute.
The Chongqing 人工智能 Institute, co-established with Peking University's School of Intelligence, leverages top academic institutions and industry leaders to cultivate local talent. Beyond building the talent foundation for the future, precise positioning for future industries is crucial, which is the key role of the Chongqing General 人工智能 Research Institute. It conducts industrial applied research focusing on three main directions: intelligent agents, embodied intelligence, and social governance.
As the core area for the "Two Institutes and One Bay," Chongqing's Nan'an District already hosts over two hundred municipal-level and above innovation platforms, constructing a full-chain support system for general 人工智能. Breakthroughs from "0 to 1" are accelerating their transformation towards the industrial end.
The "Two Institutes and One Bay" is just a microcosm. Chongqing is making every effort to build clusters of 人工智能 incubators and industrial agglomeration zones to promote the implementation of achievements and the gathering of industrial ecosystems. The deployments from the municipal plenary session send a strong signal: the wave of 人工智能 is surging forward. Chongqing will deepen and expand 人工智能 innovation applications across three dimensions—system, ecosystem, and scenarios—actively seizing new AI tracks to build momentum for the future.
In 2025, the scale of Chongqing's core 人工智能 industry grew by 23.6% year-on-year. AI empowerment drove a 19.2% increase in the city's industrial technological transformation investment, with the growth rate ranking 5th nationally. Correspondingly, there are nearly 400 enterprises in the AI industry chain, and 53 high-level scientific innovation platforms have been established. New research institutions and industrial innovation complexes are "springing up like mushrooms." Scenario applications like "AI + Enterprise Code Service" have achieved full coverage of 1.31 million enterprises, further solidifying the foundation for building a highland for 人工智能 innovation and application.
According to Chongqing's planning targets, by 2030, the scale of the core 人工智能 industry is expected to exceed 50 billion yuan. The added value of the intelligent economy's core industries is targeted to reach 4% of GDP. These core indicators point towards the "intelligent transformation" driven by 人工智能, with innovation as the core driving force.
For an intelligent terminal product, the most difficult hurdle from concept to mass production is pilot-scale production, also known as engineering trials. At the city's first AI Intelligent Terminal Pilot Verification Center located in the Mingyue Lake Science and Innovation Park, a new smart terminal product for astrophotography has entered small-batch production. It took less than a year from design drawings to trial manufacturing.
The AI Intelligent Terminal Pilot Verification Center can provide full-chain verification for the industrialization of AI intelligent terminals. New-generation AI products incubated by the Mingyue Lake Science and Innovation Base, such as AI robotic dogs and AI seasoning machines, will also enter pilot production to accelerate their market launch. The area is also building Chongqing's first AI Intelligent Terminal Shared Factory, allowing newly developed AI products to complete the entire process from concept to mass production "in one stop."
As one of the cities with the most complete range of industrial categories nationwide, Chongqing possesses both smart manufacturing scenarios derived from its industrial genes and unique scenarios like smart transportation spurred by its mountainous terrain. Chongqing's approach to AI implementation is not "finding scenarios for technology" but "attracting technology with scenarios." The "AI + Manufacturing" scenario demand list released in March of this year gathered 102 real-world demands.
With policy and demand working in tandem, the goal is to build 300 comprehensive "AI+" application scenarios by 2030. This plenary session further iterated and upgraded the "Artificial Intelligence +" action, specifying the creation of a batch of "AI+ Future Laboratories" and accelerating the construction of "full lifecycle" engineering empowerment platforms to propel the pace of 人工智能 empowering all industries.
To draw a vivid analogy for the value realization of 人工智能: technology is the hammer, application is the nail, and the scenario is the wall. Only by swinging the hammer forcefully to "drive" the nail into the wall can value be realized. The "force" in this analogy for 人工智能 is computing power. As of 2025, the computing power introduced via "Xinjiang Computing to Chongqing" exceeded 10,000 Petaflops (P), and Chongqing's "dispatchable" intelligent computing power scale surpassed 30,000 P.
According to plan, by 2027, the "Xinjiang Computing to Chongqing" computing power will reach 50,000 P. By 2030, Chongqing's "dispatchable" intelligent computing power is targeted to reach 150,000 P. Peta-scale computing power refers to small and medium-sized supercomputing and "enterprise-level" computing clusters, forming the solid foundation of computing infrastructure. Exascale (E) computing power refers to "national-level" supercomputing and top-tier AI training clusters, representing the upward development of computing capabilities. Chongqing's target for this year is to reach 14 Exaflops (E), with intelligent computing accounting for half.
In the future, "all computing will be AI." Further perfecting the computing power puzzle will bridge the final gap for AI empowerment across all industries, opening up a new competitive track.
The name "Model Power Highland" sounds futuristic, but its substance is highly capable. In the core area of the "Model Power Highland" at the Zhaomushan Lakeside Enterprise Park, within a 100,000 square meter community space, 人工智能 innovation enterprises, advantageous resources like computing power models, and service platforms for scenario testing are rapidly gathering. The supporting structure where "upstairs and downstairs form upstream and downstream" promotes the symbiotic prosperity of the industrial chain.
The "Model" in "Model Power Highland" points to model/intelligent agent technology, the "Power" represents the computing power foundation, and the "Highland" carries industrial aspirations—not a simple industrial 1+1, but an ecosystem for the integrated development of "AI + industry," becoming a "tropical rainforest" for the vigorous growth of the AI industry. By 2030, the "Model Power Highland" aims to attract and gather over 500 AI innovation enterprises.
Statistics show that a medium-sized AI enterprise consumes approximately 50 Petaflops of computing power daily, equivalent to about 600 high-end computers running at full capacity simultaneously for an entire day. The computing power demand from 500 AI enterprises gathered together is可想而知. How to supply this effectively? Chongqing has placed the answer within rows of data server racks.
At the China Mobile (Chongqing) Data Center in Liangjiang New Area, 25,000 data server racks operate day and night. As a national hub node in the Chengdu-Chongqing region of the national integrated computing power network, it has formed a total computing power scale of 939 Petaflops. This data center is just one piece of Chongqing's computing power landscape. The city has established a "one core, two zones, multiple points" computing power layout, with "Xinjiang Power to Chongqing" + "Xinjiang Computing to Chongqing" also introducing large-scale green electricity-powered computing.
From intra-city expansion to cross-regional scheduling, Chongqing is exploring innovative models of "computing-power coordination" to further solidify the foundation for "model power." Beyond this, the service matrix built by Liangjiang New Area for the "Model Power Highland," spanning from computing power algorithms to data applications, smoothens the "last mile" of the new computing power economy track.
In Chongqing's new innovation blueprint, the "Model Power Highland" is a core platform. Relying on this platform, the mountain city seeks a systematic leap in AI core competitiveness. Behind this lies the accelerating pace of a western computing power hub.
With the rapid iteration of 人工智能 development, how can Chongqing further carve out a differentiated and distinctive path? In the view of Professor Yuan Ye, Director of the Department of Business Administration and doctoral supervisor at Chongqing University of Posts and Telecommunications, several approaches are key. First, strengthen applied basic research to break through traditional disciplinary barriers and enhance the supply of foundational knowledge for "AI + all industries." Second, promote the integration of AI science and industry by conducting R&D and iteration of multi-disciplinary general-purpose AI foundation models, and building "AI+ Future Technology Laboratories" and "AI+ Future Industrial Brains" to support general scientific research tasks.
Third, mine high-value application scenarios and carry out demonstration and promotion. In the manufacturing sector, advance intelligent transformation of production processes; in the medical field, improve service supply methods driven by large models; in finance, energy, transportation, and other industries, promote the substantive embedding of 人工智能 in areas like intelligent decision-making, risk identification, and operational scheduling, using scenarios as the牵引 to accumulate experience for large-scale application. Fourth, strengthen the cultivation of interdisciplinary,复合型 talent. Encourage universities and enterprises to jointly build AI modern industrial colleges, practice bases, joint laboratories, and other industry-academia-research platforms,打造 a talent system that integrates "industry-education融合" with "science-education融汇," enhancing talent's practical transformation capabilities.
Facing the great wave of 人工智能, Chongqing possesses the advantage of its manufacturing基因 resonating with supportive policies. However, the智能经济 is advancing by leaps and bounds, and competition on new tracks waits for no one. Only by maintaining a steady, fast pace can Chongqing continuously seize opportunities. Looking further ahead, as the application boundaries of 人工智能 continuously expand, Chongqing will also迎来 windows of opportunity. Seizing these opportunities will also involve应对 challenges. This city will answer with sustained innovation, making the "intelligent transformation" visible and the future promising.
Comments