Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
You are now subscribed
Your newsletter sign-up was successful
An account already exists for this email address, please log in.
Subscribe to our newsletter
Nvidia announced today that it has invested $2 billion in Marvell Technology and entered a partnership connecting Marvell to Nvidia's AI factory and AI-RAN ecosystem through NVLink Fusion, the tech that allows third-party silicon to plug directly into Nvidia's proprietary interconnect fabric.
Marvell is one of the two dominant custom ASIC design houses alongside Broadcom. Its clients include AWS, for which it has helped develop the Trainium series of AI accelerators, as well as Microsoft and Google. These custom chips exist, in large part, to give hyperscalers an alternative to buying Nvidia GPUs, making Nvidia's investment in the company somewhat noteworthy.
Per the deal, Marvell will provide custom XPUs and NVLink Fusion-compatible scale-up networking, while Nvidia will supply Vera CPUs, ConnectX NICs, Bluefield DPUs, NVLink interconnect, and Spectrum-X switches. The two companies will also collaborate on silicon photonics and AI-RAN infrastructure for 5G and 6G networks.
Article continues below
"The inference inflection has arrived. Token generation demand is surging, and the world is racing to build AI factories," said Jensen Huang, founder and CEO of Nvidia. "Together with Marvell, we are enabling customers to leverage Nvidia's AI infrastructure ecosystem and scale to build specialized AI compute."
NVLink Fusion, first announced in May last year, enables heterogeneous AI infrastructure where non-Nvidia accelerators can communicate with Nvidia GPUs, CPUs, and networking hardware over NVLink's high-bandwidth, low-latency fabric. Platforms built through the program must include at least one Nvidia product, whether a CPU, GPU, or switch.
Marvell's contribution, meanwhile, focuses on custom XPUs and high-speed optical interconnects. The company reported $8.2 billion in revenue for its fiscal year 2026 (ended January 2026), with data center revenue accounting for more than 74% of the total. Marvell's Celestial AI acquisition late last year added photonic fabric technology to its portfolio, and this deal now places that capability inside Nvidia's ecosystem.
"By connecting Marvell's leadership in high-performance analog, optical DSP, silicon photonics and custom silicon to Nvidia's expanding AI ecosystem through NVLink Fusion, we are enabling customers to build scalable, efficient AI infrastructure," said Matt Murphy, chairman and CEO of Marvell.
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
By pulling Marvell into the NVLink Fusion ecosystem, Nvidia ensures that custom XPUs designed by Marvell remain compatible with, and dependent on, Nvidia's broader infrastructure. Every NVLink Fusion platform requires at least one Nvidia component, so Marvell-designed ASICs that use the fabric still generate Nvidia revenue.
Marvell joins an NVLink Fusion ecosystem that has grown steadily since its launch. Samsung Foundry joined in October to offer design-to-manufacturing support for NVLink-compatible custom chips. Arm entered the program in November, enabling its licensees to build CPUs with native NVLink connectivity. Nvidia rivals AMD, Intel, and Broadcom remain absent and are instead backing the open UALink standard as a competing rack-scale interconnect.
Luke James is a freelance writer and journalist. Although his background is in legal, he has a personal interest in all things tech, especially hardware and microelectronics, and anything regulatory.
