- DeepSeek and Huawei collaborate to develop tools for Ascend AI chips
- DeepSeek open-sources programming libraries for Huawei's Ascend platform
- The partnership advances a supernode using 128 Ascend 950 chips
Chinese AI startup DeepSeek has teamed up with Huawei Technologies to develop programming tools optimised for Huawei's Ascend AI chips, as Chinese technology companies deepen efforts to build alternatives to Nvidia's ecosystem, Reuters reported.
DeepSeek said in a post on its official WeChat account on Wednesday that it was open-sourcing programming infrastructure for Huawei's Ascend platform, including libraries designed for computing and communication.
The move comes about two weeks after Huawei unveiled its next-generation AI processors and supernode computing systems, saying it expected the systems to see wider use for AI model training next year.
ALSO READ: China Probes AI Startups DeepSeek, Moonshot Over Claude Data Routing, Breaches: Report
According to DeepSeek, Huawei provided full support for the development of the programming infrastructure. The two companies have also jointly advanced a supernode solution built around 128 Ascend 950 chips, with work focused on optimising both computing and communication.
DeepSeek also highlighted the role of TileLang, a high-level open-source programming language designed for AI chips. The company said the language can improve development efficiency and simplify code logic.
DeepSeek said building a new generation of independent and self-controlled GPU software ecosystems would require a high-level programming language that is easy to use while still allowing developers to access the full performance of the underlying hardware.
It described TileLang as being designed for that purpose, offering what it called a simpler programming model than Nvidia's CUDA.
The partnership underscores the growing importance of software infrastructure alongside AI chips. While Huawei has been developing its own Ascend hardware, compatible programming tools are crucial for developers seeking to build and run AI models without relying on Nvidia's widely used CUDA ecosystem.
ALSO READ: DeepSeek Ends Weekend Peak Pricing For API Users From Today
For China's AI industry, the collaboration could help strengthen an alternative hardware-software stack around domestic chips, even as companies continue to navigate restrictions on access to advanced US-made semiconductors and related technologies.
Essential Business Intelligence, Sharp Market Insights, Practical Personal Finance Advice, Daily Fuel, Gold and Silver Prices and Latest Stories — On NDTV Profit.