NVIDIA is set to launch its next-generation GB300 AI server lineup at the GPU Technology Conference (GTC) in March 2025, with Taiwan-based Foxconn continuing its role as the company’s largest server supplier. According to Economic Daily News, the development and design phase for the GB300 servers is already underway.
This new AI server lineup is a significant step for NVIDIA in maintaining its dominance in the AI market. Foxconn, which has partnered with NVIDIA since 2017, will again be crucial in the production of the GB300, handling key components such as chip modules and assembly, while also contributing to the development of liquid cooling systems.
The GB300 servers are expected to incorporate a 100% liquid cooling solution, potentially eliminating traditional fans. This shift is likely to benefit Foxconn, as its vertical integration allows it to provide up to 90% of the server’s components, excluding the GPU and CPU.
NVIDIA’s GB200 servers, the predecessor to the GB300, have already been integral to the company’s AI-driven workloads, and with the GB300, NVIDIA aims to continue delivering high-performance, energy-efficient computing solutions for data centers and AI applications. With NVIDIA’s AI market growth, the launch of the GB300 is set to further solidify the company’s position, while providing sustained demand for Foxconn’s manufacturing capabilities.
Related topics:
NVIDIA and TSMC Showcase Silicon Photonics; Samsung Accelerates Development
NVIDIA Investigates AI Chip Sales to China Following U.S. Request