What are the key benefits of Alibaba Cloud's M890 supernode?

The M890 supernode allows enterprises to rent high-performance computing resources with 64 accelerators and low-latency interconnects. This reduces infrastructure costs and accelerates the deployment of large-scale AI models.

What AI workloads is the M890 supernode designed for?

The solution is optimized for inference of mixture-of-experts models with up to 10 trillion parameters. It suits complex AI tasks requiring high computational power.

Which AI models are already running on the M890 supernode?

Models like Kimi K3 and Qwen3.8-Max are already deployed on the M890 supernode, providing enterprise services through Alibaba Cloud.

← All news

Artificial intelligence

Alibaba Cloud Launches M890 AI Supernode in China

August 12, 2026

Photo: TechNode

Quick answer

Alibaba Cloud has launched the M890 AI supernode for enterprise customers in China, enabling processing of AI models with up to 10 trillion parameters.

Alibaba Cloud has begun commercial deployment of the Lingjun Zhenwu M890 AI supernode across China, with Ulanqab as the first region. Enterprise customers can now rent computing resources featuring 64 accelerators and high-speed interconnects.

The solution is designed for inference of mixture-of-experts models with up to 10 trillion parameters. This enables processing of complex AI workloads without the need to build private data centers, reducing operational costs and accelerating technology adoption.

Models such as Kimi K3 and Qwen3.8-Max are already running on the M890 supernode, delivering enterprise-grade services. Alibaba Cloud continues to expand its cloud infrastructure to meet growing AI computing demands.

Common questions

What are the key benefits of Alibaba Cloud's M890 supernode?: The M890 supernode allows enterprises to rent high-performance computing resources with 64 accelerators and low-latency interconnects. This reduces infrastructure costs and accelerates the deployment of large-scale AI models.
What AI workloads is the M890 supernode designed for?: The solution is optimized for inference of mixture-of-experts models with up to 10 trillion parameters. It suits complex AI tasks requiring high computational power.
Which AI models are already running on the M890 supernode?: Models like Kimi K3 and Qwen3.8-Max are already deployed on the M890 supernode, providing enterprise services through Alibaba Cloud.

Dzen feed: /feed/dzen.xml · RSS: /feed.xml