China expands AI ecosystem through infrastructure and clean energy

China is expanding its artificial intelligence capabilities by focusing on the entire technology stack and leveraging its clean energy surplus. Efforts to achieve self-sufficiency are moving beyond models and processors into software, networking, and infrastructure. For instance, DeepSeek has open-sourced programming infrastructure for Huawei’s Ascend platform, and the two companies have collaborated on a 128-chip Ascend 950 supernode.

To support these computing needs, Beijing is implementing the ‘East Data, West Computing’ program. This strategy aims to relocate computing resources to western regions where solar, wind, and hydropower are abundant, rather than transporting electricity to crowded eastern cities. The government has established eight national computing hubs and ten data-center clusters to facilitate this. Under this plan, new data centers are required to source over 80% of their electricity from clean sources.

While these developments could strengthen China’s position in cost-sensitive overseas markets, challenges remain. Constraints in advanced semiconductors, memory, and software maturity persist. Additionally, the rapid expansion of renewable energy faces bottlenecks in transmission infrastructure, raising questions about how effectively power can be mobilized for AI workloads.