China is accelerating global compute deployment and scaling AI infrastructure through international partnerships to circumvent U.S. export restrictions.

Chinese firms are pushing their domestic computing ecosystem overseas despite U.S. export controls that restrict access to high-end AI chips. The People's Republic of China faces an uneven compute crunch—shortages concentrated in frontier hardware even as aggregate computing capacity expands. This dynamic enables Chinese firms to export compute without first achieving hardware parity with foreign competitors.

Beijing is laying groundwork to internationalize its computing ecosystem before resolving technological and supply constraints. Chinese academics, policy elites, and industry figures have begun calling for domestic "compute going global" (算力出海). Shenzhen's AI action plan, released in early September, similarly calls for domestic foundational hardware to expand abroad as part of an integrated Chinese AI ecosystem. Chinese sources suggest this push reflects a shift away from waiting for supply constraints to erode; instead, Chinese firms increasingly view internationalization as a way to open new markets, bolster demand for Chinese technology, and establish Chinese hardware, software, and technical standards globally. While exporting compute entails near-term opportunity costs at home, Chinese actors appear willing to bear them for long-term technological and commercial gains.

Chinese firms distinguish between two models of overseas expansion. Physical deployment represents one approach, in which hardware itself is shipped and installed abroad. More mature is "services crossing the border" (服务出境), where computing hardware remains in the PRC while domestic data centers process AI workloads for foreign customers and export resulting tokens and services. Even when hardware remains physically in the PRC, compute exported as a service represents a deliberate decision to devote computational resources to foreign rather than domestic demand, since computing capacity is inherently rivalrous.

Cross-border compute exports have developed into an organized industry push. At the 2026 World Artificial Intelligence Conference, one exhibitor featured a dedicated English-language "Token going global" (Token出海) section demonstrating Chinese computing power packaged into standardized token services for export, attracting particular interest from Middle Eastern attendees. In 2025, the China Academy of Information and Communications Technology (CAICT)—the think tank under the Ministry of Industry and Information Technology—launched a dedicated "Computing-Power Token Going Global Ecosystem Initiative" (算力词元 (Token) 出海生态计划) in collaboration with industry partners. Its organizers aim to move beyond experimental projects toward "standardized, efficient, and sustainable" (规范、高效、可持续) overseas expansion and argue this could hedge external risks while expanding services trading.

Chinese firms are actively laying groundwork for physical deployment of advanced domestic AI hardware abroad. Huawei has drawn up plans to export Ascend-series chips, including the Ascend 950DT and 950PR AI accelerator, to South Korea and has signed distributor agreements with local partners including Hansol PNS and SK Shieldus. The company has unveiled the Atlas 960E SuperPoD to address large-scale AI infrastructure demand overseas. Other Chinese companies have announced plans to deploy a "domestic computing power cluster" (出海国产算力集群) in Saudi Arabia. Malaysia is considering adopting Huawei's 910C Ascend in its sovereign AI ambitions.

Chinese academic and policy elites have endorsed this wider push. Wang Zhongyuan of Fudan University has argued that the PRC should assist Global South countries by constructing local computing infrastructure, jointly building data centers, and exporting "computing services" (算力服务). Song Guoyou, deputy director of Fudan University's Center for American Studies, explicitly calls for Chinese firms to deploy "computing clusters" (算力集群) across regions of the Global South.

A growing number of Chinese industry and policy figures explicitly describe internationalization as the next stage for the domestic compute industry. Li Wei, deputy director of CAICT's Cloud Computing and Digitalization Research Institute, has called for accelerating overseas expansion and integrating the domestic computing-power industry into the "global computing-power industrial system" (全球算力产业体系). Wang Chao, executive director of CAS Star—a deep-tech venture capital firm founded by Chinese Academy of Sciences researchers—has similarly argued that domestic compute will serve not only domestic demand but also overseas expansion.

Chinese companies articulate corresponding ambitions. Huawei has actively touted Ascend and its Kunpeng computing platform as creating "a new choice for the world" (为世界构建了新的选择). iFlytek has partnered with Huawei Ascend to "cooperate deeply on overseas large-model deployments" (在海外主权大模型上深度合作) and "plant flags and seize positions around the world" (在全球插红旗、抢阵地). Other firms describe computing power going global as the "industry's second growth curve" and argue that overseas expansion "is no longer optional, but imperative."

Southeast Asia has become a particularly important potential market. A CAICT report constructed a market-entry evaluation framework for overseas expansion and identified opportunities for Chinese computing-power providers in emerging markets, with particular focus on Southeast Asia's surging demand. One Chinese AI infrastructure company projects compute demand in Southeast Asia will exceed 3GW in 2026 and 6GW in 2028, arguing that Chinese firms should leverage accumulated technological capabilities to "rapidly enter overseas computing-power center markets" (快速切入海外算力中心市场) and "capitalize on the supply gap" (把握 … 供给缺口) in Southeast Asia's high-end data center market. Sugon, another major Chinese computing-infrastructure provider, has established an ASEAN business headquarters in Nanning, Guangxi, offering a complete solution combining Sugon servers with high-end Hygon DCU compute.

Chinese hardware providers achieve pricing advantages by undercutting foreign competitors. Estimates suggest that for equivalent computing services yielding similar inference performance, Chinese compute providers can charge just one-fifth to one-seventh as much as comparable overseas services. This price advantage derives partly from lower electricity costs in the PRC and vertical integration of large Chinese telecommunications operators, most of which own both computing and network infrastructure.