This article is based on ideas originally published by VoxEU – Centre for Economic Policy Research (CEPR) and has been independently rewritten and extended by The Economy editorial team. While inspired by the original analysis, the content presented here reflects a broader interpretation and additional commentary. The views expressed do not necessarily represent those of VoxEU or CEPR.
Yazhang No. 1 Equipped with High‑Performance Servers and an Eco‑Friendly Cooling System Part of the Dongxisushan Project, Implemented in the Western Plateau Region Eight Computing Hubs to Be Built Nationwide to Realize a Digitally Connected Economy
China’s large-scale computing network project, designed to connect the entire country, has now expanded to the Tibetan Plateau. A newly built, large-scale AI-focused computing center in the Yarlung Zangbo River Basin of Shannan City, Tibet Autonomous Region, has officially begun operations. With this, China has expanded its supercomputing infrastructure beyond traditional northwestern regions, such as Ningxia and Gansu, into Tibet, introducing a high-efficiency data center model that leverages the natural environment and renewable energy sources of the western highlands.
According to the South China Morning Post (SCMP) on June 29, Tibet’s first large-scale AI-dedicated computing center, named Yazhang No. 1, held its completion ceremony and officially began operations on June 18 in the Yarlung Zangbo River Basin, Shannan City, Tibet Autonomous Region. The SCMP highlighted that this project marks the first implementation of China’s national “Eastern Data, Western Computing” (Dongshu Xisuan) strategy in the western plateau region, signaling Tibet’s entry into a fast-growth phase of digital economic development.
Launched in 2022, the Dongshu Xisuan strategy aims to shift the massive data processing demands of China’s eastern coastal cities to energy-rich western regions. While previous expansions reached places like Ningxia Hui Autonomous Region and Gansu Province, this is the first time the initiative has extended into Tibet.
A joint investment by Tibet Yazhang Computing Technology Company and Naidong District of Shannan City, the project has deployed 256 high-performance servers in its initial phase, delivering a total computing capacity of 2,000 petaflops (one petaflop equals one quadrillion calculations per second).
Yazhang No. 1, developed in collaboration between the local government and Tibet Yarlung Zangbo Computing Technology Company, offers 2,000 petaflops of processing power during its initial stage, with more than 256 advanced computing servers installed. To minimize its environmental impact, the facility utilizes a high-efficiency cooling system that combines solar power and waste heat recovery, resulting in a Power Usage Effectiveness (PUE) of less than 1.3. This represents a 40% improvement in energy efficiency compared to conventional data centers.
Last year, the National Data Administration (NDA) of China announced plans to leverage the Eastern Data, Western Computing (Dongshu Xisuan) initiative as the backbone for building a nationwide large-scale computing power network. The NDA emphasized that although the computing power gap between the eastern and western regions had narrowed in recent years, significant imbalances remained. The project was specifically designed to address this disparity by efficiently linking the massive data processing demands of the eastern region with the abundant power and natural resources available in the west, ultimately enhancing China’s overall digital competitiveness.
A key pillar of the Dongshu Xisuan strategy is the establishment of eight national computing hubs across the country. Four of these are located in economically advanced eastern regions: the Beijing-Tianjin-Hebei (Jingjinji) cluster, the Yangtze River Delta (encompassing Shanghai, Jiangsu, Zhejiang, and Anhui), the Greater Bay Area (comprising Guangdong, Hong Kong, and Macau), and the Chengdu-Chongqing (Chengyu) economic circle. These regions host large-scale industrial complexes and technology companies, serving as the central engines of China’s digital economy. They are home to massive data centers, supercomputing facilities, and intelligent computing infrastructure that support critical digital services, such as AI and big data analytics.
The remaining four hubs, Inner Mongolia, Guizhou, Gansu, and Ningxia, are situated in western China. With access to vast land, low energy costs, and abundant electricity from both renewable and conventional sources, these regions are being developed to house large-scale data centers and serve as the nation’s primary providers of computing power. These western hubs help distribute the data processing load of the East, improve energy efficiency, and stimulate local economic growth and the development of emerging industries.
China’s National Development and Reform Commission, NDA, and the Ministry of Industry and Information Technology are jointly supporting the initiative. The goal is for these hubs to become the country’s most significant contributors to national computing capacity by 2025. This effort aims not only to bridge the digital divide between regions but also to upgrade China’s nationwide data infrastructure.
Since being designated a National Computing Power Hub in May 2021, Guizhou Province has rapidly transformed into the digital economy center of Western China. Once known mainly as a large-scale data storage location, Guizhou has evolved into a high-performance computing engine and a key node in China’s Eastern Data, Western Computing (Dongshu Xisuan) initiative.
By the end of 2024, Guizhou’s total computing power reached 57 exaflops (EFLOPS), more than doubling within a short period. Notably, over 90% of this capacity is now used for intelligent computing tasks. With significant infrastructure, including the Huawei Ascend Computing Center and other cluster-based facilities, Guizhou has positioned itself as one of the top intelligent computing hubs in China.
To build on this momentum, the provincial government has introduced 26 new policy incentives aimed at attracting enterprises from major economic zones like Guangdong, Sichuan, and Beijing. These include financial incentives and support programs designed to foster a robust digital industry ecosystem. Efforts are concentrated in areas like Gui’an New District, where a large-scale data center cluster continues to expand, integrating both resource allocation and computing capacity for maximum efficiency.
By the end of 2024, Guizhou accounted for 23% of China’s total computing power, with its computing services generating transaction volumes of USD 1.56 billion.
Guizhou’s computing capacity is already playing a visible role in the entertainment and media industries, especially in film and animation rendering. The Gui’an Supercomputing Center has provided computing services for over 50 significant projects, including visual effects for “The Wandering Earth,” “The Three-Body Problem,” and “Ne Zha 2.”
Despite its growth, Guizhou faces challenges such as achieving technological self-reliance, acquiring skilled talent, and coordinating with other regions. To address this, the province has launched scientist workstations and university-industry research labs aimed at cultivating innovation talent. By the end of 2025, China’s largest computing cluster is scheduled for completion in Guizhou, positioning the province as a strategic hub in the global digital economy.
Every statistic about public trust in science reveals less a knowledge gap than an abyss of distrust, a chasm that reason alone cannot bridge. Despite decades of data, surveys, and outreach campaigns, roughly three-quarters of Americans—76% in the latest Pew Research Center poll—report having at least some confidence in scientists to act in the public’s best interest; yet, more than one in four express little to no trust at all.
This article was independently developed by The Economy editorial team and draws on original analysis published by East Asia Forum. The content has been substantially rewritten, expanded, and reframed for broader context and relevance. All views expressed are solely those of the author and do not represent the official position of East Asia Forum or its contributors.
This article was independently developed by The Economy editorial team and draws on original analysis published by East Asia Forum. The content has been substantially rewritten, expanded, and reframed for broader context and relevance. All views expressed are solely those of the author and do not represent the official position of East Asia Forum or its contributors.
This article is based on ideas originally published by VoxEU – Centre for Economic Policy Research (CEPR) and has been independently rewritten and extended by The Economy editorial team. While inspired by the original analysis, the content presented here reflects a broader interpretation and additional commentary. The views expressed do not necessarily represent those of VoxEU or CEPR.
All OTT News
AGI: The Key Variable That Will Define Future Revenue Models OpenAI Can Restrict Microsoft's Access If It Declares AGI Clash Over Clause Allowing Early Termination of Microsoft's Exclusive Tech Rights
The once-powerful alliance between OpenAI and Microsoft, which has driven the global AI revolution, is now showing serious signs of fracture. What began as a groundbreaking partnership, pairing OpenAI’s cutting-edge AI models with Microsoft’s cloud computing power, has evolved into a high-stakes dispute that could reshape the future of artificial intelligence.
At the heart of the conflict is a controversial contract clause tied to the achievement of Artificial General Intelligence (AGI), AI capable of performing intellectual tasks at or beyond human levels. The clause allows OpenAI to revoke Microsoft’s exclusive access to its technology if OpenAI itself declares that it has achieved AGI.
When the agreement was made, this seemed like a distant possibility. But OpenAI’s rapid technological breakthroughs have turned this clause into a very real point of contention. What’s more, the dispute has evolved far beyond technical definitions, morphing into a broader battle over commercial control, profit-sharing, independence, and the strategic direction of AI.
The world is now watching closely as one of the most important partnerships in the AI industry teeters on the edge of collapse.
The conflict first came to light in late June when reports revealed that OpenAI and Microsoft were locked in a tense standoff over the AGI clause in their contract. Microsoft, OpenAI’s largest backer with an investment exceeding USD 13 billion since 2019, holds a 49% stake in OpenAI’s for-profit entity. Their agreement stipulates that should OpenAI declare the achievement of AGI, Microsoft’s exclusive technical usage rights would be terminated.
Initially dismissed as a theoretical safeguard, this clause has become a genuine flashpoint as OpenAI accelerates toward what it believes could soon be AGI. OpenAI CEO Sam Altman has confidently declared, “We know how to build AGI,” signaling the company’s belief that such technology is within reach in the near future. OpenAI has also started securing direct partnerships with major tech players like Apple and Salesforce, signaling its move to operate more independently of Microsoft.
Microsoft, however, remains skeptical. CEO Satya Nadella has dismissed the AGI label as “a subjective, self-congratulatory concept with no concrete substance.” He has gone so far as to suggest that a valid AGI claim should only be made if it leads to monumental, measurable impacts, such as a 10% increase in global GDP growth.
These divergent views are not merely philosophical; they underpin a fierce power struggle. OpenAI seeks to renegotiate its contract terms to reduce Microsoft’s revenue share and secure the right to independently commercialize high-value AI products beyond Microsoft’s Azure cloud. Conversely, Microsoft is demanding the complete removal of the AGI clause and wants to extend its exclusivity rights beyond 2030, aiming to prevent OpenAI from gaining independence through an AGI declaration.
Microsoft fears that despite its massive financial commitments and integration of OpenAI’s models into Azure, Copilot, and Bing Chat, it could be cut off from AGI, the most valuable AI milestone, if OpenAI exercises the AGI clause. On the other side, OpenAI has expressed growing frustration over Microsoft’s overbearing control of its AI products, computing infrastructure, and intellectual property, particularly given restrictions that force OpenAI to distribute its services solely through the Azure platform.
This legal and technical dispute has escalated into a full-blown commercial war with significant implications for both companies and the broader AI industry.
OpenAI is aggressively pursuing a transition from a nonprofit to a fully for-profit corporation, a move it argues is necessary to secure the tens of billions in funding needed to sustain rapid AI development. However, this pivot has drawn heavy criticism. Elon Musk, one of OpenAI’s original co-founders, filed a lawsuit last year accusing OpenAI and Sam Altman of violating the company’s founding principles and breaking agreements made with early investors by shifting toward a profit-driven model. Figures like Nobel Prize-winning physicist Geoffrey Hinton, Meta CEO Mark Zuckerberg, and several civil society organizations have also formally opposed the shift, arguing that OpenAI’s nonprofit mission has been compromised.
Despite the backlash, OpenAI has reignited efforts to restructure as a for-profit entity but requires Microsoft’s approval to proceed. Negotiations, however, have reached an impasse. Microsoft is demanding a 35% ownership stake in the new for-profit structure as a condition for its consent, an ask that OpenAI views as excessive and monopolistic.
In response, OpenAI is reportedly considering a bold countermeasure: filing an antitrust complaint against Microsoft, accusing the company of abusing its dominant position to demand unreasonable control over OpenAI’s future.
The fallout from this power struggle extends far beyond the two companies. Should negotiations fail, OpenAI’s IPO ambitions could collapse, and Microsoft’s AI-driven growth strategy, anchored in products like Copilot and Azure, would suffer a massive blow. OpenAI would gain its long-sought independence but face the monumental challenge of rebuilding its entire cloud infrastructure from scratch.
Meanwhile, competitors like Google, Amazon, and Meta stand ready to capitalize on any rift, potentially accelerating a major realignment in the AI and cloud computing markets. Analysts warn that the AI industry could enter an even more cutthroat phase of competition, with market dynamics shifting dramatically if the OpenAI-Microsoft partnership dissolves.
This high-stakes confrontation is no longer just about an AGI clause buried in a contract, it’s a fight over who controls the future of AI itself. Whether OpenAI and Microsoft reconcile or part ways, the outcome will have profound consequences for the development, commercialization, and governance of artificial intelligence globally.
The result could determine whether AGI, should it be achieved, remains in the hands of a single corporate giant, becomes part of an open and competitive ecosystem, or splinters among rival tech companies locked in an arms race for dominance.
What began as one of the most celebrated partnerships in the AI era now serves as a cautionary tale of how ambition, commercial interests, and the pursuit of technological supremacy can strain even the most formidable alliances. As the world watches closely, the future of AI, and possibly the future of how humanity interacts with intelligent machines, hangs in the balance.
