Nvidia Invests $3.5 Billion in MediaTek
Nvidia has invested $3.5 billion in MediaTek as the companies deepen their collaboration across artificial intelligence infrastructure, local computing and automotive technology. The investment came through convertible bonds issued by the Taiwanese semiconductor designer.
Nvidia Expands Its MediaTek Partnership
The deal strengthens a partnership that already covers several areas of AI computing. Moreover, MediaTek will adopt Nvidia’s NVLink Fusion platform to develop custom AI accelerators.
Through the platform, hyperscalers, cloud providers and AI developers can build custom XPUs that connect with Nvidia’s rack-scale AI systems. Consequently, customers can develop specialized processors while retaining compatibility with Nvidia’s broader computing infrastructure.
The companies said their expanded work will span three major areas. These include AI infrastructure, local AI computing and automotive platforms.
Focus Shifts Across AI Computing
For data centers, MediaTek will use NVLink Fusion to help customers integrate custom AI infrastructure with Nvidia systems. Meanwhile, the companies will continue developing RTX Spark and DGX Spark chips for consumer PCs, AI developer systems and enterprise workstations.
The partnership also extends into vehicles. Specifically, Nvidia and MediaTek will develop platforms for AI-powered, software-defined vehicles as automotive computing advances.
In addition, the companies are building on earlier work involving Nvidia’s GB10 Grace Blackwell Superchip. MediaTek contributed to the system, which combines Nvidia’s Blackwell GPU with its Grace CPU through NVLink-C2C connectivity.
Nvidia Targets a Broader AI Ecosystem
The investment also gives Nvidia a stronger connection to MediaTek’s custom silicon capabilities. Therefore, the companies can address workloads that require specialized processors alongside Nvidia’s accelerated computing platforms.
The announcement comes as technology companies increasingly develop custom AI silicon. However, Nvidia’s strategy allows it to support that shift while keeping its NVLink ecosystem central to large-scale AI infrastructure.
