ByteDance is reportedly deploying thousands of Nvidia AI chips in Malaysia, highlighting how Chinese tech firms are navigating U.S. export controls to expand global computing power.
 |
| ByteDance’s reported plan to deploy Nvidia Blackwell chips in Malaysia highlights how Chinese tech firms are building AI capacity overseas amid tightening U.S. semiconductor controls. Image: CH |
BEIJING, China, March 13, 2026:
ByteDance, the Chinese technology company that owns TikTok, is reportedly expanding its artificial intelligence computing capacity outside China using advanced chips from Nvidia, a move that reflects the increasingly complex global competition for AI infrastructure.
According to a report by The Wall Street Journal, ByteDance is working with Southeast Asian cloud provider Aolani Cloud to deploy roughly 500 Nvidia Blackwell computing systems in Malaysia. The project could involve around 36,000 of Nvidia’s B200 AI chips, among the most powerful processors designed for training and running large artificial intelligence models.
The reported deployment highlights how Chinese technology firms are increasingly building AI computing capacity overseas as geopolitical tensions reshape global semiconductor supply chains.
In recent years, the United States has imposed export restrictions on advanced AI chips destined for China, citing concerns that powerful processors could be used to accelerate military and surveillance technologies.
As a result, Chinese companies have sought alternative strategies to secure computing resources needed to train large AI models and operate advanced recommendation algorithms. Establishing computing infrastructure in third countries is emerging as one possible workaround.
Malaysia and other Southeast Asian countries are becoming attractive locations for such projects due to their growing data center infrastructure, relatively open technology policies and proximity to major Asian markets.
The reported scale of the deployment—36,000 advanced chips—underscores how AI development is increasingly dependent on vast amounts of computing power.
Companies building large language models, recommendation engines and generative AI systems require massive clusters of graphics processing units (GPUs) and specialized AI accelerators. Nvidia’s newest processors, including the Blackwell architecture, are designed specifically for these workloads.
For ByteDance, which operates one of the world’s most data-intensive social media platforms, expanding AI computing capacity is critical. Algorithms power TikTok’s recommendation engine, advertising tools and emerging generative AI features.
The project also reflects the broader geopolitical struggle surrounding semiconductor technology.
Washington’s restrictions on advanced chips are part of a wider strategy to slow China’s progress in cutting-edge computing and artificial intelligence. But technology companies are increasingly adapting their infrastructure strategies in response.
By building computing clusters outside mainland China, firms may be able to access high-performance chips while still serving global operations and AI research.
The development also highlights Southeast Asia’s rising role in the global technology ecosystem.
Countries such as Malaysia are rapidly expanding data center capacity to support cloud computing and AI workloads. Investments from global technology companies are transforming the region into an important hub for digital infrastructure.
For firms like ByteDance, this shift provides both technological and strategic advantages: access to advanced chips, proximity to Asian markets and diversification of computing resources.
However, such arrangements could attract further scrutiny from policymakers in Washington.
If U.S. authorities believe that advanced chips deployed abroad are indirectly supporting Chinese AI development, additional export restrictions or regulatory oversight could follow.
The situation illustrates a broader challenge facing the global tech industry: as artificial intelligence becomes a strategic technology, decisions about where computing infrastructure is built are increasingly shaped not just by business considerations, but by geopolitics.