Nvidia has invested $2 billion in Marvell Technology as part of a push to make it easier for customers to deploy custom artificial intelligence chips designed by Marvell alongside Nvidia’s networking equipment and processors.
Marvell shares rose around 7% following the announcement, whilst Nvidia gained about 2.7%. The deal is aimed at ensuring Nvidia remains central to meeting the growing computing demands of AI, even as some companies shift toward custom-built processors over its higher-cost chips.
Analysts noted that the partnership gives Nvidia access to Marvell’s semi-custom silicon and advanced optical interconnect technologies, helping scale data centre AI systems where bandwidth and power efficiency are critical constraints. It also expands Nvidia’s ecosystem to include more specialised silicon, reinforcing its role as a key platform for a wider range of AI workloads.
The collaboration is expected to reduce friction in integrating third-party AI chips within Nvidia-led data centre environments, allowing the company to maintain its dominance whilst broadening the overall AI semiconductor landscape.
Both firms will work on advanced networking solutions for AI, with a focus on optical interconnects and silicon photonics to enable faster and more energy-efficient data transmission. Marvell will provide custom chips and networking solutions compatible with Nvidia’s NVLink Fusion, whilst Nvidia will contribute supporting technologies such as CPUs, network interface cards, and interconnects.
Meanwhile, major technology companies including Alphabet and Meta are projected to spend at least $630 billion on AI infrastructure this year, boosting demand for server and networking chips and supporting companies like Marvell.
Marvell has also projected nearly 40% revenue growth, with expectations to approach $15 billion in fiscal 2028.
Click here for more on Business
