How Powerful are Huawei’s New SuperPoDs and SuperClusters?
Huawei uses its annual HUAWEI CONNECT 2025 event in Shanghai to showcase new large-scale computing systems designed to power the next generation of AI.
Eric Xu, Huawei's Deputy Chairman of the Board and Rotating Chairman, delivers a keynote speech entitled Groundbreaking SuperPoD Interconnect: Leading a New Paradigm for AI Infrastructure, where he announces the launch of the Atlas SuperPoD and SuperCluster portfolio.
"Computing power is – and will continue to be – key to AI. This is especially true in China," Eric tells the audience.
“Chips are the building blocks of computing power. And at Huawei, Ascend chips are the foundation of our AI computing strategy.”
He explains that Huawei aims to build SuperPoDs and SuperClusters with semiconductor manufacturing process nodes that are available to the Chinese mainland, ensuring sustainable long-term supply of computing power for data centre environments.
The scale of SuperPoDs and SuperClusters
A SuperPoD is a single logical machine made up of multiple physical machines that work together to learn, think and reason as one system.
At the event, Huawei introduces the Atlas 950 SuperPoD, with 8,192 Ascend Neural Processing Units (NPUs), and the Atlas 960 SuperPoD, with 15,488 Ascend NPUs. NPUs are specialised processors designed to accelerate AI workloads such as machine learning and deep learning.
These SuperPoDs provide high levels of computing performance across four critical metrics: number of NPUs, total computing power, memory capacity and interconnect bandwidth.
Eric explains that, based on publicly available product roadmaps from other companies, the Atlas 950 and 960 SuperPoDs are the most powerful systems of their kind globally and will retain that position for years to come.
The company also launches the Atlas 950 SuperCluster, with over 500,000 Ascend NPUs, and the Atlas 960 SuperCluster, with over one million Ascend NPUs. A SuperCluster is a large-scale computing system made up of multiple SuperPoDs, enabling exponential increases in scale and performance.
Eric highlights that these clusters are set to outperform all other available computing clusters in the market, making them highly relevant for hyperscale data centres.
"With the world's most powerful SuperPoDs and SuperClusters," says Xu, "Huawei has what it takes to provide abundant computing power for ongoing, rapid advancements in AI, both now and in the future."
Expanding into general-purpose computing
In addition to AI-optimised systems, Huawei introduces the TaiShan 950 SuperPoD, the world's first general-purpose computing SuperPoD. General-purpose computing systems are designed to handle a wide range of workloads beyond AI training, including enterprise databases and transactional systems.
The TaiShan 950 SuperPoD works in combination with Huawei's distributed GaussDB database and offers a potential alternative to traditional mainframes, mid-range servers and Oracle's Exadata database systems. This positions the TaiShan 950 as a flexible option for data centres that require both AI and enterprise-grade computing.
Overcoming interconnect bottlenecks
One of the main challenges for large-scale AI computing infrastructure is interconnect technology. Traditional optical and copper cables face physical limitations when linking vast numbers of chips and SuperPoDs across long distances while maintaining high-speed, low-latency connections.
Huawei uses its three decades of expertise in connectivity to address this bottleneck. By combining this experience with system-level innovation, the company develops a new interconnect protocol called UnifiedBus. This protocol underpins the performance of the new SuperPoDs and ensures reliable and scalable communication across the infrastructure.
At the event, Eric also reveals UnifiedBus 2.0 technical specifications and calls on industry partners to support adoption. The aim is to establish an open UnifiedBus ecosystem where more vendors develop products and components that can integrate with the protocol.
"SuperPoDs and SuperClusters powered by UnifiedBus are our answer to surging demand for computing, both today and tomorrow," concludes Eric. "Our goal is to keep pushing advancements in AI to create greater value."
