Huawei will once again introduce some dynamic Ascend AI chip solutions at the 2026 WAIC (World Artificial Intelligence Conference). The company has announced that it’ll bring more innovative technologies to the stage, contributing to China’s AI goals.
WAIC is China’s flagship state-supported AI summit. It is one of the world’s largest annual events, specifically themed for artificial intelligence. It brings together many global tech executives, like top scientists, investors, and government officials.
As per Huawei China, the company will unveil new Ascend AI chip solutions like superPoDs, industry practices, and case studies at WAIC 2026. It’ll happen from July 17 to 20.
SuperPoD refers to a large-scale computational unit made of many physical machines working together as a single logical machine and ensures task completion efficiency.
Inputs further revealed that the Chinese tech giant will focus on four key areas. These segments include the industry’s largest supernode – Atlas 950 SuperPoD, an air-cooled supernode cluster, open-source AI foundational software, and supernode deployment at industry scale.
Let’s check these four segments one by one:
(Image Credits: Weibo)
1. Atlas 950 SuperPoD – Huawei will bring the industry’s largest supernode to obtain system-level innovative breakthroughs through globally unified memory.
From basic components and protocol algorithms to optoelectronic technologies, it will address all kinds of AI bottlenecks and accelerate the training + inference deployment of trillion-parameter large-scale models.
2. Atlas 850E Air-Cooled Supernode – It will be a real-world demonstration unit that will let you enjoy supernode technology in an air-cooled server room. The solution will use upgraded VCE cooling tech to offer stability, high throughput, and a low-latency experience for Agentic inference.
3. Ascend AI open-source foundational software – The tech is compatible with third-party training acceleration libraries, inference engines, and several programming languages, continuously optimizing the developer experience, moving from easy-to-use functionality to easy-to-operate. It also improves training + inference efficiency of mainstream large-scale models.
4. Supernode Deployment at Industry Scale – The new AI solution will deeply engage with the core business scenarios in the Internet, Telecom Operator, Finance, Power, and other customer sectors.
