Strategic collaboration and a $2 billion investment accelerate AI factory innovation, advanced networking, and telecom transformation.
Santa Clara, CA — March 31, 2026 — NVIDIA today announced a major expansion of its AI ecosystem through a strategic partnership with Marvell Technology, Inc. (NASDAQ: MRVL), aimed at delivering unprecedented flexibility, scalability, and performance for next-generation AI infrastructure.
As part of the collaboration, NVIDIA has invested $2 billion in Marvell, reinforcing a shared vision to power the global surge in AI-driven computing. The partnership integrates Marvell into the NVIDIA AI factory and AI-RAN ecosystem via NVLink Fusion™, enabling customers to build highly customizable and efficient AI systems.
Powering the AI Factory Era With NVLink Fusion
At the heart of this collaboration lies NVLink Fusion, NVIDIA’s advanced rack-scale platform designed to support semi-custom AI infrastructure. By combining Marvell’s custom XPUs and scale-up networking capabilities with NVIDIA’s full-stack ecosystem, customers gain the ability to design tailored AI architectures without compromising compatibility.
NVIDIA will provide key technologies, including its Vera CPU, ConnectX® NICs, BlueField® DPUs, NVLink interconnect, and Spectrum-X™ switches—delivering a powerful, end-to-end AI infrastructure stack.
This integration ensures seamless interoperability across NVIDIA GPUs, networking, and storage platforms, while giving enterprises the freedom to innovate with heterogeneous computing models.
Driving the Next Wave of AI and Telecom Innovation
The partnership goes beyond data centers. NVIDIA and Marvell are set to transform telecommunications networks into AI-powered infrastructure, leveraging NVIDIA Aerial AI-RAN for 5G and emerging 6G networks.
Additionally, the companies will collaborate on silicon photonics and advanced optical interconnect technologies, paving the way for faster, more efficient data movement—critical for scaling AI workloads globally.
Leadership Insights on the AI Inflection Point
“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 scale rapidly and build specialized AI compute using our powerful ecosystem.”
“Our expanded partnership with NVIDIA highlights the growing importance of high-speed connectivity and accelerated infrastructure in AI,” said Matt Murphy, chairman and CEO of Marvell. “By combining our strengths in custom silicon, optical DSP, and silicon photonics with NVIDIA’s AI leadership, we are empowering customers to build scalable and efficient AI systems.”
Enabling a New Standard for AI Infrastructure
With AI workloads becoming increasingly complex and data-intensive, this collaboration positions NVIDIA and Marvell at the forefront of the industry’s evolution. The integration of customizable XPUs, high-performance networking, and advanced interconnects ensures enterprises can meet rising demands for AI training, inference, and real-time processing.
By bridging semiconductor innovation with AI ecosystem scale, the partnership delivers a future-ready foundation for cloud providers, enterprises, and telecom operators worldwide.
About Marvell Technology, Inc.
Marvell is a leading provider of data infrastructure semiconductor solutions, empowering the world’s top technology companies for over three decades. Through deep collaboration and cutting-edge innovation, Marvell enables the movement, storage, processing, and security of global data across cloud, enterprise, and carrier networks.
About NVIDIA
NVIDIA is the global leader in artificial intelligence and accelerated computing, driving breakthroughs across industries with its powerful GPUs, AI platforms, and full-stack computing ecosystem.
Forward-Looking Statements
This press release contains forward-looking statements, including expectations regarding the benefits, impact, and performance of the NVIDIA–Marvell partnership, as well as broader trends in AI infrastructure and telecommunications. These statements are subject to risks and uncertainties that may cause actual results to differ materially. Factors include market conditions, technological developments, supply chain dependencies, regulatory considerations, and competitive dynamics. Both companies undertake no obligation to update forward-looking statements except as required by law.
