Observed Signal · Sep 16, 2026 · Strategy Update · Source: Hello China Tech · Impact: 4/5 · Sentiment: Positive
Huawei adopts three-tier AI model strategy: infrastructure, in-house, hybrid
Huawei's supervisory board chairman Guo Ping outlined a three-tier AI model strategy based on a Q&A transcript published on September 10. For ICT and computing, Huawei aims to be like Nvidia, providing infrastructure (Ascend, Kunpeng) for any model. For devices and intelligent driving, in-house models are preferred. Huawei Cloud must develop its own models to stay competitive. The strategy is constrained by limited Ascend compute, creating a resource allocation hierarchy among business units. Guo also cited DeepSeek's founder on eroding Nvidia's CUDA moat, discussed Ascend's progress, Yinwang's ownership, and China's compute gaps, while the 2025 annual report shows strong automotive growth and cloud revenue decline.
Huawei's AI strategy is significant for the AdTech industry as it shapes the AI infrastructure landscape, potentially affecting how large-scale AI models are deployed and monetized across platforms, which is relevant for advertising technology relying on AI-driven targeting and optimization.
Track Huawei Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Huawei's supervisory board chairman Guo Ping disclosed a three-tier AI model strategy in a Q&A with new employees.
- For ICT and computing, Huawei aims to be like Nvidia, supporting all models on Ascend/Kunpeng infrastructure.
- Huawei prefers in-house models for devices and intelligent driving, while Huawei Cloud must develop its own to stay competitive.
- Huawei's 2025 annual report shows ICT revenue at Rmb 375bn, consumer at Rmb 344.5bn, and automotive solutions at Rmb 45bn (up 72.1%).
- Huawei Cloud's external revenue shrank 3.5% to Rmb 32.2bn, while total revenue including internal transactions was Rmb 72.1bn.
Connected Companies & Entities
5 Entities mapped“Huawei's three-tier AI model strategy and resource allocation....”
“Huawei aims to be like Nvidia in supporting customer models....”
“Guo cited DeepSeek founder Liang Wenfeng's transcript on AI moats....”
“Guo noted Anthropic as a company that made breakthroughs in large models....”
“Guo noted Moonshot AI's Kimi as a breakthrough in large models....”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Samsung to invest $1B in AI infrastructure firm Helix
Samsung Electronics and five affiliates will invest a combined $1 billion in Helix Digital Infrastructure, an AI infrastructure company launched by KKR and backed by Nvidia. Samsung Electronics contributes $500 million, with the rest from Samsung C&T, Samsung SDS, Samsung SDI, Samsung Life Insurance, and Samsung Fire & Marine Insurance. Helix, led by former AWS CEO Adam Selipsky, focuses on hyperscale data centers, power generation, transmission, and fiber-optic networks. The investment adds to over $10 billion already committed by other investors including KKR, Kuwait Investment Authority, Nvidia, and Vistra. The move allows Samsung to leverage its semiconductor, cooling, data center construction, and battery capabilities to expand in the AI infrastructure market.
GLM-5.3 Sparse Attention Impact on DRAM Memory TAM
This article analyzes the impact of sparse attention mechanisms, specifically DeepSeek Sparse Attention (DSA) used in Z.ai's GLM-5.3 model, on the total addressable market (TAM) for DRAM memory, including HBM and NAND. It explains that while sparse attention reduces KV cache memory and bandwidth during the attention operation, it does not reduce overall memory capacity requirements because the top-k selection still requires full context in HBM. The article discusses system optimizations like HiSparse, which offloads KV cache to host DRAM to overcome capacity bottlenecks. It also provides detailed performance and cost comparisons for serving GLM-5.3 on different hardware (GB200, GB300, MI355X) using inference engines like Dynamo-SGLang, Dynamo-TRT-LLM, and ATOM, highlighting cost-efficiency and interactivity trade-offs. The analysis includes a deep dive into GLM-5's architecture, including the lightning indexer, MLA configuration, and post-training pipeline.
Alibaba's T-Head AI Chips: Cloud Customers or Qwen Training?
Alibaba announced at its Apsara conference that its new Zhenwu V900 AI chip will enter mass production and go on sale in Q1 2027, two quarters earlier than planned. This follows Huawei's announcement of its Ascend 960DT chip being ready in Q1 2027. Both companies face high demand and limited supply for their chips. IDC data shows Nvidia holds 55% of China's server AI accelerator shipments, Huawei 20%, and T-Head 7%. Alibaba plans to train Qwen models with 5-10 trillion parameters, but has not disclosed which chips will be used, raising concerns about competition between internal model training and paying cloud customers for scarce chip capacity.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.