Taiwan-based electronics manufacturer Foxconn continues to be the largest supplier of NVIDIA’s servers, with the firm central to NVIDIA’s plans to launch its next-generation GB300 AI server lineup at the GTC (GPU Technology Conference) in March 2025, according to the Taiwanese media outlet Economic Daily News. The supply chain has entered the development and design phase for the GB300 AI server, the report said.
Similar to Apple’s strategy of launching a new iPhone each year, NVIDIA CEO Jensen Huang had previously announced that the company will unveil a new generation of AI products annually.
Why it matters: The launch of the GB300 AI server lineup marks a critical step for NVIDIA in maintaining a competitive edge in the AI market, while also benefiting the company’s largest server supplier, Foxconn, through sustained demand for AI-related products.
Details: Foxconn secured the largest share of GB300 orders due to its highly vertically integrated resources, the Economic Daily News report says. NVIDIA and Foxconn have had a close partnership since 2017, working together on the development of AI servers from the first generation to the current GB200 models.
- In addition to handling the chip modules and assembly for the AI servers, Foxconn has also invested in areas such as liquid cooling systems and connectors, and is currently conducting related validations.
- Industry analysis suggests that NVIDIA may eliminate fans in the GB300 servers and adopt a 100% liquid cooling solution. NVIDIA’s component procurement rights are expected to be led by the assembly manufacturers, which will significantly benefit Foxconn.
- The group’s highly vertical integration of resources is one of the key factors in its close cooperation with customers, said Foxconn Chairman Liu Yangwei. Taking the GB200 model as an example, Foxconn can provide about 80% to 90% of the components, excluding the GPU and CPU, Liu added.
Context: The NVIDIA GB200 is a series of advanced AI servers designed to deliver high-performance computing for data centers and AI-driven workloads. It integrates powerful GPUs and CPUs to accelerate machine learning, data analysis, and other AI applications. The GB200 series is a key component of NVIDIA’s strategy to provide scalable, energy-efficient solutions for industries requiring large-scale computing power.
- NVIDIA’s revenue for the third quarter was $35.1 billion, up 17% from the previous quarter and 94% year-on-year. During the same quarter, NVIDIA’s net profit reached $19.3 billion, a 16% increase from the previous quarter and a 109% rise compared to last year.
Related
