News
Foxconn targets L11 and L12 rack solutions for NVIDIA Vera Rubin platform
Sunday, March 1, 2026 at 09:58 AM
Foxconn is expanding its role in the AI supply chain by targeting L11 server systems and L12 rack solutions for NVIDIA's next-generation Vera Rubin platform. The company aims to showcase integrated solutions involving cold-plate manufacturing and liquid cooling at the upcoming GTC conference. While Foxconn's vertical integration from L1 to L12 is expected to drive revenue, industry watchers note that the high cost of components like GPUs and HBM could pressure gross margins if not managed through diverse transaction models.
Context
Foxconn is preparing a major push for NVIDIA’s Vera Rubin platform, targeting high-value L11 and L12 rack solutions at the upcoming GTC conference this March. Following NVIDIA’s late-February shipment of initial samples, Foxconn expects to commence mass production in the second half of 2025. By moving beyond component manufacturing into full-rack integration—featuring advanced liquid cooling and proprietary sensor technology—Foxconn aims to secure a dominant position in the next-generation AI supercomputer supply chain.
This strategic shift is expected to significantly boost profitability for Foxconn Industrial Internet (FII), which projects its AI server revenue will triple in 2025. To mitigate the margin pressure from expensive GPUs and HBM, Foxconn is leveraging a "turnkey" vertical integration model. This approach combines hardware assembly with internal software calibration and mechanical engineering, allowing the company to command margins that exceed its historical average as it scales for the Vera Rubin rollout.
Related Companies
Nvidia
NVDA
Foxconn
2317
Foxconn Industrial Internet
601138