Nvidia's $7 billion acquisition of Mellanox has evolved into a cornerstone of its AI strategy, with networking revenue surging to $1.2 billion in Q4 2025—up 68% year-over-year. The integration of InfiniBand and Ethernet technologies has become a critical enabler for high-performance data centers, fueling demand across cloud and enterprise sectors.
- Nvidia’s Mellanox acquisition cost $7 billion in 2020
- Networking revenue reached $1.2 billion in Q4 2025, up 68% YoY
- Networking now accounts for 14% of Nvidia’s total revenue
- Nvidia holds over 40% market share in high-end AI cluster networking
- NVDA stock is up 27% YTD, outperforming XLK ETF’s 11% gain
- Hyperscalers like AWS, Microsoft Azure, and Google Cloud are major adopters
Nvidia's strategic pivot following its $7 billion acquisition of Mellanox in 2020 has yielded transformative results. The integration of Mellanox’s high-speed networking technologies—InfiniBand and SmartNICs—into Nvidia’s AI infrastructure stack has become a key growth engine. In the fourth quarter of 2025, networking revenue reached $1.2 billion, marking a 68% year-over-year increase and accounting for 14% of total revenue, up from 8% in 2023. This expansion underscores the shift from standalone GPU sales to bundled AI solutions. The momentum is driven by demand from hyperscalers like Microsoft Azure, Amazon Web Services, and Google Cloud, which now rely on Nvidia’s end-to-end AI platforms. These platforms combine GPUs with Mellanox-powered interconnects to optimize data flow across thousands of servers, reducing latency and boosting training efficiency. The result is a tighter ecosystem that locks in enterprise customers and increases lifetime value. Market analysts note that Nvidia’s networking segment has outperformed both Intel and AMD in the data center interconnect space. While Intel’s Ethernet and AMD’s Xilinx-based offerings remain niche, Nvidia’s unified architecture has captured over 40% of the high-end AI cluster networking market. This dominance has contributed to a 27% rise in NVDA’s stock price year-to-date, outpacing the broader XLK ETF’s 11% gain. The long-term implications extend beyond Nvidia’s balance sheet. As AI workloads grow in complexity, the importance of low-latency networking is becoming a decisive factor in infrastructure decisions. This has intensified competition in the semiconductor sector, with Intel and AMD accelerating their own interconnect roadmaps to counter Nvidia’s momentum.