Huawei: Ascend Has Passed the Ecosystem Tipping Point, Has the Ability and Opportunity to Build a New AI Ecosystem
Summary
On September 19, Zhu Zhaosheng, head of Huawei's Computing Strategy and Business Development Department, said Ascend has crossed the ecosystem tipping point and is fully capable of building a new AI ecosystem in the Agentic era. His evidence: the Ascend CANN open-source community averages over 5,200 monthly active users and has been China's most active open-source community since June, with non-Huawei developers now outnumbering Huawei developers and more than 30,000 lines of code merged daily; over 40 large language and multimodal models have been pretrained on Ascend, making it the only AI compute platform in China supporting pretraining, and it now connects to mainstream frameworks such as PyTorch.
Bubble analysis
This is about the software ecosystem and developer activity around Huawei's Ascend chips — progress in China's domestic compute supply chain — but it discloses no investment, revenue, order or capacity figures. A healthier ecosystem could mean more future Chinese AI compute spending, yet on its own it neither supports nor refutes the AI bubble thesis; it sits several steps away from the question of whether too much money is being spent.
Related signals
linked by shared tags
- CITIC Securities: Ascend 960 Launch Moved Up, China's Domestic Compute Supply Chain Set to Benefit#huawei#ascend#chips
- Huawei's Zhu Zhaosheng: Ascend Has Crossed the Ecosystem Inflection Point and Is Fully Capable of Building a New AI Ecosystem in the Agentic Era#huawei#ascend#chips
- Interview with Huawei's Wang Tao: Building AI Compute Competitiveness with Super Nodes and Clusters#huawei#chips
- China Telecom Open-Sources Xing4.0-29B-A4B, Its First Fully Domestic-Stack Lightweight Agent LLM#ascend#open-source
- Huawei's Wang Tao: Huawei Aims to Build an AI Compute Foundation — Doing One Chip Well Is Far From Enough#huawei#chips