DeepSeek and Huawei Team Up to Build AI Chip Tools as China Seeks Alternatives to Nvidia
September 30, 2026
Chinese artificial intelligence company DeepSeek has partnered with Huawei Technologies to develop programming tools designed specifically for Huawei’s Ascend AI chips, marking another step in China’s efforts to strengthen its domestic AI hardware and software ecosystem.
DeepSeek said it is open-sourcing programming infrastructure for Huawei’s Ascend platform, including computing and communication libraries. The development could make it easier for AI developers to build and optimise applications on Huawei hardware rather than relying entirely on Nvidia’s widely used software ecosystem.
DeepSeek Focuses on Huawei Ascend Chips
According to DeepSeek, Huawei provided full support during the development of the new programming infrastructure. The companies have also worked together on a supernode solution using 128 Ascend 950 chips, with efforts focused on improving both computing and communication performance.
The announcement comes shortly after Huawei unveiled its next-generation AI processors and supernode systems. Huawei has said it expects its AI computing systems to see wider use for model training in the coming year.
TileLang Becomes an Important Part of the Project
One of the key technologies highlighted by DeepSeek is TileLang, a high-level open-source programming language designed for AI chips.
DeepSeek says TileLang can simplify programming and improve development efficiency while allowing developers to take advantage of the underlying hardware's capabilities. The company also described its programming model as simpler than Nvidia's CUDA.
The software development is significant because AI chips need more than powerful hardware. Developers also require programming frameworks, libraries and tools that allow AI models to run efficiently on those processors.
China Pushes for a More Independent AI Ecosystem
The partnership comes as Chinese technology companies continue developing alternatives to foreign AI hardware and software. Restrictions on advanced semiconductor exports have increased the importance of domestic computing technologies for China's AI industry.
By developing software that works closely with Huawei's Ascend processors, DeepSeek is helping address one of the major challenges facing alternative AI chips: creating a strong software ecosystem around the hardware.
The move does not mean Nvidia's technology has immediately been replaced. Nvidia's CUDA ecosystem remains widely used by AI developers globally. Instead, the DeepSeek-Huawei collaboration represents an effort to build additional options for AI computing within China's technology sector.
What This Could Mean for the AI Chip Race
The development highlights how competition in artificial intelligence is increasingly extending beyond chip manufacturing.
Powerful processors need compatible programming languages, libraries and developer tools to become practical at scale. DeepSeek and Huawei's work therefore puts software infrastructure at the centre of China's attempt to expand the use of domestically developed AI chips.
As AI model development continues to grow, the ability to combine hardware with an effective software ecosystem could become increasingly important for companies and countries seeking greater control over their computing infrastructure.
Conclusion
DeepSeek's partnership with Huawei marks a significant development in China's domestic AI technology push. By open-sourcing tools for Huawei's Ascend chips and working on a 128-chip Ascend 950 supernode, the companies are seeking to strengthen an alternative AI computing ecosystem.
The long-term impact will depend on how widely developers adopt the tools and how effectively Huawei's hardware can support increasingly demanding AI workloads.