Nvidia invests $2 billion in Marvell Technology to make custom AI chips more accessible to customers: Read CEO Jensen Huang's statement

Nvidia has invested $2 billion in Marvell Technology as part of a strategic partnership. With this deal, the chip giant wants to make it easier for its customers to integrate Marvell's custom AI chips with Nvidia's networking hardware and central processors. The deal centres on Nvidia's NVLink Fusion platform, a rack-scale system that allows customers to build semi-custom AI infrastructure within the Nvidia ecosystem. Under the arrangement, Marvell will contribute custom accelerators, known as XPUs, and NVLink Fusion-compatible scale-up networking. Nvidia will supply supporting technologies, including its Vera CPU, ConnectX network interface cards, BlueField data processing units, NVLink interconnects, and Spectrum-X switches.The two companies will also collaborate on silicon photonics technology, which enables high-speed, energy-efficient data transmission through optical interconnects, and work together to bring AI capabilities to telecommunications networks through Nvidia's Aerial AI-RAN platform for 5G and 6G infrastructure.The partnership comes as Nvidia works to remain central to the AI infrastructure market, even as companies increasingly explore custom processors as alternatives to its standard chips. In a statement, Nvidia founder and CEO Jensen Huang said, "The inference inflection has arrived. Token generation demand is surging, and the world is racing to build AI factories. Together with Marvell, we are enabling customers to leverage NVIDIA's AI infrastructure ecosystem and scale to build specialized AI compute."Marvell chairman and CEO Matt Murphy said the expanded partnership reflects "the growing importance of high-speed connectivity, optical interconnect and accelerated infrastructure in scaling AI," adding that connecting Marvell's capabilities in analogue, optical DSP, silicon photonics, and custom silicon to Nvidia's ecosystem through NVLink Fusion would allow customers "to build scalable, efficient AI infrastructure."The deal comes as major technology companies, including Google-parent Alphabet and Facebook-parent Meta, are expected to collectively spend at least $630 billion on AI infrastructure this year, driving demand for chips used in servers and networking equipment. Marvell has said it expects its revenue to grow nearly 40% and approach $15 billion in fiscal 2028.