Huawei launches AI data centre infrastructure platform
Author: Joe Peck
Chinese multinational technology company Huawei has introduced a new full-stack data infrastructure platform designed for AI data centres and large-scale enterprise AI deployments.
The announcement was made during the Huawei Innovative Data Infrastructure (IDI) Forum 2026 on 21 May in Paris, France, where Yuan Yuan, Vice President of Huawei and President of the company’s Data Storage Product Line, outlined the growing role of AI infrastructure in enterprise operations.
According to Huawei, increasing adoption of AI agents and AI applications is driving significant growth in enterprise data processing and token consumption, requiring organisations to redesign traditional IT infrastructure around AI workloads.
With this in mind, the company says its platform has been developed to support AI data lakes, inference systems, model deployment, agent frameworks, and data resilience.
Platform targets large-scale AI workloads
Huawei says the infrastructure platform combines storage, data management, model deployment, and AI orchestration technologies intended for enterprise and hyperscale AI environments.
The company introduced updates across several areas of its AI infrastructure portfolio, including:
• OceanStor Pacific Scale-Out Storage for high-density data storage
• DME Omni-Dataverse unified data management platform
• Context Memory Storage for AI inference environments
• ModelEngine deployment and resource scheduling platform
• Nexent AI agent framework
• AI-focused data resilience and protection technologies
Firstly, Huawei says its OceanStor Pacific storage platform can deliver 11PB of capacity within a 2U chassis, enabling massive data storage at “optimal total cost of ownership” (TCO), while DME Omni-Dataverse is designed to support multimodal and cross-site data management.
The company also introduced its Context Memory Storage system, which it says is designed for large-scale inference clusters and can reportedly reduce time to first token by 90%.
AI infrastructure evolves beyond compute alone
Huawei states that AI infrastructure planning increasingly requires integration between storage, compute, models, and data management systems, rather than focusing solely on GPU capacity.
The company says its ModelEngine platform supports model deployment and compute resource scheduling, including partitioning resources across multiple workloads.
Huawei also introduced the Nexent agent platform, which is designed to allow AI agents to be generated through natural language interactions.
“AI is unlocking new opportunities for the IT industry,” says Yuan. “The next chapter of AI is data.
“Committed to technological innovation in data storage, Huawei will accumulate the experience of industrial AI adoption and work closely with the entire industry to help customers accelerate their journey into the intelligent era.”
The company says the platform also includes technologies designed to address AI-related security risks including ransomware, data tampering, and data poisoning attacks.
For more from Huawei, click here.
