
Huawei Atlas · Supernode / Cluster
Atlas 900 AI Cluster
Composed of thousands of Ascend training processors, the Atlas 900 AI cluster combines HCCS, PCIe and RoCE high-speed interconnects through Huawei cluster communication libraries and scheduling for image, speech and NLP model training, research and autonomous driving.
Key Highlights
- Industry-leading scale with thousands of interconnected Ascend processors
- excellent cluster network integrating HCCS, PCIe and RoCE.
Technical Specifications
| Interconnect | HCCS, PCIe, RoCE |
|---|---|
| Processors | 数千颗昇腾(Ascend)训练处理器 |
| NPU count, compute power and memory | 官网未公开 |
Typical Use Cases
Image, speech and NLP model trainingscientific researchautonomous driving development.
翻译稿;官网页面仅定性描述,未公开处理器数量、AI算力(PFLOPS)、内存等量化参数;img1: main; 无公开技术白皮书链接。
Related Products

Huawei Atlas
Atlas 200I DK A2 Developer Kit
Easy-to-use AI development kit for fast innovation View details →
Huawei Atlas
Atlas 200I A2 Acceleration Module
Edge AI accelerator module embedded in smart devices, robots and drones View details →
Huawei Atlas
Atlas 300I A2 Inference Card
High-performance AI inference card integrated into servers View details →
Huawei Atlas