Nvidia is investing $3.5 billion in MediaTek through bonds convertible into MediaTek shares, as the two companies expand their AI chip partnership announced today in the US [1, 2, 3, 4]. MediaTek will adopt Nvidia’s NVLink Fusion and NVHBM technologies to design AI chips integrated into Nvidia-based data centers, enabling data center owners and AI companies to create custom chips that work seamlessly within Nvidia’s AI ecosystem [1, 2, 3, 4].
The collaboration also targets AI infrastructure for personal computers and AI-powered vehicles, with plans to develop next-generation chips for Nvidia's RTX Spark and DGX Spark platforms [4]. MediaTek’s automotive Dimensity Auto systems already incorporate Nvidia technologies and operate alongside Nvidia's DRIVE AGX platform for intelligent vehicle cockpits [4].
Nvidia CEO Jensen Huang, who was born in Taiwan, highlighted the partnership’s scope: "AI is transforming every computing platform — from the world’s largest AI factories to the PC and the car" [4]. The deal strengthens the ties between Nvidia and MediaTek, which is expanding beyond smartphones into custom AI chips [1, 2, 4].
The investment comes amid moves by cloud and AI giants like Amazon, Google, Microsoft, OpenAI, and Anthropic to develop their own chips and reduce reliance on Nvidia GPUs [1, 2, 3]. Amazon separately announced a partnership to deploy an additional 2 million Nvidia components and use Nvidia’s connection technology alongside its in-house chips [1, 2, 3].
Dion Harris, Nvidia senior director of HPC and AI hyperscaler infrastructure solutions, said, "Nvidia is an AI infrastructure company. We expanded beyond pure computing chips years ago. Basically, every cloud, every model builder is deploying our platform in some shape, form, or fashion" [3].
Nvidia has previously invested in companies integrating into its AI ecosystem, leading to some investor concerns about circular investments, which Nvidia disputes [1, 2, 3].
The expanded partnership aims to provide customers more custom AI chip options fully compatible with Nvidia’s dominant AI data center platform [1, 2, 3, 4]. The companies signed the deal and announced the investment and collaboration plans on August 31, 2026 [1, 2, 3, 4].
