[REVISION HISTORY]

Huawei AI and connectivity strategic development

Updated 4 times since CLSTR started tracking revisions of this situation.

What changed

2026-09-18 21:02 UTC → 2026-09-20 21:22 UTC · added ~~removed~~

Huawei is advancing its strategic focus on artificial intelligence, connectivity, and computing infrastructure. At the M360 ASEAN 2026 event, the company presented strategies for AI-driven digital transformation in the ASEAN region, highlighting the ‘Huawei Agentic MBB’ and the ‘OpenAN’ open-source architecture. Huawei also released a joint report with GSMA Intelligence regarding the monetization of AI economic models. In its ‘Intelligent World 2035: From Vision to Action’ report, Huawei outlined ten key industry directions, including the development of Artificial General Intelligence (AGI) and advanced computing clusters. The company predicts that by 2035, global annual token consumption will increase 100,000-fold, driven by AI agents. To support this, Huawei unveiled a grid-interactive AI data center (AIDC) solution designed to maximize ‘tokens per watt’ while managing high-density power and cooling requirements. During the ‘Act Today. Shape Tomorrow’ launch event, Rotating Chairman David Wang identified agentic AI as a ‘critical variable’ in the transition of AI from the digital realm into the physical world. Leadership has clarified that the core strategy centers on connectivity and computing strengths. Rather than developing solely proprietary models, Huawei aims to support customers in building large-scale models using its Ascend and Kunpeng platforms. To address semiconductor manufacturing constraints, the company intends to utilize architectural innovation and integrated design-to-manufacturing advantages. Additionally, Huawei is expanding its service capabilities through the ‘Jingwei Plan,’ which seeks to establish an on-site service network across 300 cities in China by collaborating with over 15,000 engineering partners. At HUAWEI CONNECT 2026 in Shanghai, the company expanded its infrastructure to support the ‘agentic AI’ era. New hardware includes the OceanStor M900 Context Memory Storage, which provides a PB-scale shared memory space to address bottlenecks in multi-turn inference. To facilitate large-scale AI clusters, Huawei debuted the UnifiedBus architecture, capable of connecting up to one million NPUs.

Versions

Only revisions since CLSTR began indexing content versions appear here. Select a version to see what changed compared to the one before it.