Nvidia Invests $3.5 Billion in MediaTek as Big Tech Builds Its Own AI Chips

Nvidia is investing $3.5 billion in Taiwanese chipmaker MediaTek, deepening a partnership that spans AI data centers, consumer PCs and automotive computing as major AI companies and cloud providers push to build more of their own AI chips.

The investment, announced Monday, comes in the form of convertible bonds issued by MediaTek, according to a joint statement from the two companies. As part of the expanded collaboration, MediaTek will adopt Nvidia’s NVLink Fusion platform, giving the chipmaker’s customers a prevalidated path to design custom AI accelerators, often called XPUs, that plug directly into Nvidia’s rack-scale data center systems.

A Hedge Against Custom Silicon

The deal arrives as Amazon, Google, Microsoft, OpenAI and Anthropic all invest in developing their own AI chips to cut their reliance on Nvidia’s GPUs. Rather than fight that shift head-on, Nvidia is positioning itself as the infrastructure underneath it: MediaTek’s custom chip designs will still depend on Nvidia’s interconnects, memory technology and rack-scale systems to reach production.

“Nvidia is an AI infrastructure company,” Dion Harris, Nvidia’s senior director of HPC and AI hyperscaler infrastructure solutions, told reporters on a call Monday. “We expanded beyond pure computing chips years ago.”

“Basically, every cloud, every model builder is deploying our platform in some shape, form, or fashion,” Harris told TechCrunch. “So by MediaTek being able to offer this extension to its customers, it allows them to standardize on the rack-scale infrastructure across their AI factories… and also deploy their [custom chips] right alongside those using the same standard platform.”

The NVLink Fusion platform bundles several pieces of Nvidia technology that a chip designer would otherwise have to build or license separately: the NVLink Fusion chiplet, which connects custom chips to Nvidia’s NVLink scale-up fabric; NVLink-C2C, a high-bandwidth link between chips; and NVHBM, Nvidia’s approach to customized high-bandwidth memory. According to the companies’ announcement, the goal is to let customers focus on the parts of a chip that differentiate their product while relying on Nvidia and MediaTek for the packaging, connectivity and manufacturing work needed to get a custom accelerator into a working data center.

Three Areas of Collaboration

Nvidia and MediaTek described the expanded partnership as covering three areas. In AI infrastructure, MediaTek will use NVLink Fusion to help hyperscalers and AI labs build custom accelerators for Nvidia-connected data centers. In local AI computing, the companies will keep working together on Nvidia’s RTX Spark and DGX Spark chips, which power AI-focused desktop and developer PCs that pair Nvidia GPUs with MediaTek system-on-chip designs. In automotive, the two companies will continue building platforms for AI-powered, software-defined vehicles, extending an existing arrangement in which MediaTek’s automotive chips use Nvidia RTX graphics for vehicle cockpit displays and work alongside Nvidia’s Drive AGX platform for autonomous driving.

“AI is transforming every computing platform, from the world’s largest AI factories to the PC and the car,” Jensen Huang, founder and CEO of Nvidia, said in the companies’ statement. “MediaTek is one of the world’s great semiconductor companies, with exceptional expertise in system-on-chip design, connectivity, leading performance and power efficiency. Together, we’re building platforms that bring NVIDIA accelerated computing to new markets and give customers the freedom to create differentiated AI systems at enormous scale.”

“MediaTek and NVIDIA share a vision for making advanced AI computing pervasive across the technology landscape,” said Rick Tsai, vice chairman and CEO of MediaTek. “NVIDIA’s investment strengthens a collaboration that spans cloud AI infrastructure, local AI computing and automotive in the era of physical AI.”

Part of a Wider Pattern

The MediaTek deal follows a similar move Nvidia made with Amazon Web Services last week: AWS agreed to deploy an additional 2 million Nvidia GPUs across its infrastructure and integrate NVLink Fusion with its own custom Trainium chips, though without a direct Nvidia investment attached. Speaking with Bloomberg Television about the MediaTek deal, Huang described it as an expansion of an already large relationship. “We’ve already had a big partnership with MediaTek. Today, we are going to make it a lot bigger,” he said, adding that the two companies’ engineering roadmap now extends for a decade.

MediaTek has been building out its own custom data center chip business, telling investors in June that it expects the unit to generate $2 billion in revenue in 2026. According to Invezz, the company is aiming for roughly 15 percent of an $80 billion custom AI chip market next year, and has separately struck a partnership with Google. The Nvidia deal was also described by Invezz as Nvidia’s largest direct investment outside the United States to date.

Nvidia’s growing pattern of putting money directly into partners across its own supply chain, a pattern that also shows up in Lambda’s recent billion-dollar debt raise to buy Nvidia hardware, has drawn scrutiny from investors worried about circular financing propping up AI demand. Huang rejected that framing when asked about the MediaTek agreement. “This is not circular because obviously they do their own business and we do our own business, and MediaTek is already incredibly profitable, incredibly successful,” he said.