Nvidia's networking business has officially emerged as the company's second major growth engine, standing tall alongside its famous GPUs.
The latest financial results are nothing short of remarkable. In the fourth quarter of fiscal year 2026, Nvidia's networking revenue hit an astounding $10.98 billion, a 263% surge compared to the same period last year. This isn't just a momentary spike; it's a powerful trend that solidifies networking's crucial role in Nvidia’s dominance of the data center market.
So, what's fueling this incredible expansion? It's not a single factor, but rather a perfect storm of strategic moves and market forces. Let’s look at the three primary drivers.
First, there's the unprecedented AI factory boom. Tech giants known as hyperscalers—like Google, Microsoft, and Meta—are investing hundreds of billions of dollars to build massive data centers. These AI factories need tens of thousands of GPUs to function, but those GPUs are useless unless they can communicate with each other instantly. That's where Nvidia's networking solutions, like InfiniBand and Spectrum-X, come in. They act as the super-fast nervous system connecting everything, and as capex for AI soars, so does the demand for this essential hardware.
Second, this success was strategically engineered over many years. A pivotal moment was Nvidia's acquisition of networking specialist Mellanox in 2020. This move gave Nvidia the foundational technology it needed to build a world-class networking portfolio. Since then, the company has skillfully cultivated an ecosystem around it, partnering with major players like Cisco and Dell. This 'full-stack' approach, offering both the compute engine (GPUs) and the interconnects, creates a deeply integrated platform that is difficult for competitors to replicate.
Third, geopolitics has played an unexpected part. U.S. export restrictions have limited the sale of Nvidia's most advanced GPUs to China. This has concentrated Nvidia's growth in other global markets, where it can sell its complete, high-value systems without restriction. In a way, the policy has strengthened the business case for its all-in-one solution outside of China, further boosting networking sales.
While the growth is impressive, it's important to maintain perspective. Some online commentary has suggested Nvidia's quarterly networking revenue has surpassed Cisco's annual sales. This is not accurate. Nvidia's latest quarter is about 39% of Cisco's full-year networking product revenue. It's a testament to Nvidia's rapid ascent, but the facts matter.
In summary, the narrative of networking as Nvidia's 'second pillar' is firmly rooted in reality, driven by a long-term vision, massive AI investment, and a changing global landscape.
- Glossary
- Hyperscaler: A large-scale cloud service provider that operates massive data centers around the world, such as Google, Amazon Web Services, and Microsoft Azure.
- InfiniBand / Spectrum-X: High-performance networking technologies from Nvidia. InfiniBand is traditionally used in supercomputing, while Spectrum-X is an Ethernet-based solution optimized for AI workloads.
- Capex (Capital Expenditure): Money a company spends to buy, maintain, or improve its fixed assets, such as buildings, vehicles, equipment, or land.
