Nvidia is building an IP licensing empire on the back of NVLink
Nvidia is turning its NVLink interconnect technology into a licensing business, allowing rival chipmakers to build custom AI processors that still plug into Nvidia's ecosystem. On Monday, MediaTek became the latest company to adopt NVLink Fusion for its datacentre XPU chips, joining Amazon, Fujitsu, Qualcomm, Arm and Marvell, in a move that shows Nvidia remains central to the AI hardware market even as customers build their own chips to reduce reliance on its GPUs.
As part of the MediaTek deal, Nvidia is investing $3.5 billion in convertible bonds issued by the Taiwanese chip designer, while continuing to license MediaTek's own designs for future DGX and RTX Spark systems. NVLink Fusion, launched last year, lets rival XPUs tap into Nvidia's high-speed chip-to-chip and rack-scale networking, and because Nvidia's MGX rack designs are part of the Open Compute Project, partners can slot their own compute blades into existing NVL72 rack designs. This cuts engineering overheads for competitors while still generating revenue for Nvidia through NVSwitch sales and licensing fees, with Amazon's upcoming Trainium4 accelerators, for example, set to replace their in-house interconnect with NVLink Fusion.
- MediaTek adopts Nvidia's NVLink Fusion for its new datacentre AI chips
- Nvidia invests $3.5 billion in MediaTek convertible bonds as part of deal
- Amazon, Qualcomm, Arm and others also licensing Nvidia's interconnect tech
New here? Start with this
AI datacentres use powerful chips to train and run artificial intelligence systems. Nvidia is a leading supplier of these chips, while companies such as Amazon and MediaTek are also developing their own processors to meet their computing needs.
NVLink is Nvidia’s technology for connecting chips so they can exchange data quickly, including across large server racks. Its Fusion version lets other chipmakers connect their processors to Nvidia-based systems, and use compatible rack designs.
This could help rivals build custom processors with less engineering work while keeping them connected to Nvidia’s hardware and services. For Nvidia, licensing the technology and selling related equipment could create income even when customers use chips they designed themselves.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Licensing NVLink could make it easier for other chipmakers to offer AI processors that work with Nvidia’s widely used systems, reducing engineering costs and giving customers more choice in how they build datacentres. Nvidia can argue that shared interconnects and rack designs improve compatibility and efficiency, while licensing revenue helps fund further investment in the technology.
The case against
The deals could deepen Nvidia’s influence even as customers try to reduce their dependence on its GPUs: rival processors may still rely on Nvidia’s interconnects, switches and rack designs. A critic could argue that this risks making Nvidia a gatekeeper for competing hardware, potentially limiting customers’ freedom to switch suppliers and giving Nvidia commercial leverage over the wider AI ecosystem.