NVIDIA will invest $2 billion (approximately 3 trillion won) in Marvell Technology. This investment aims to enable chips designed by Marvell, a fabless company, to be utilized within NVIDIA’s network, central processing units (CPUs), and other elements of the artificial intelligence (AI) infrastructure ecosystem.


NVIDIA Headquarters. Photo by EPA Yonhap News

NVIDIA Headquarters. Photo by EPA Yonhap News

View original image

On March 31 (local time), NVIDIA and Marvell announced that they had formed a strategic partnership to build AI data center infrastructure. The two companies will combine NVIDIA’s NVLink Fusion networking technology with Marvell’s XPU-based AI semiconductors to develop custom AI systems. They also plan to expand collaboration in related areas such as silicon photonics (optical semiconductors), high-speed optical interconnect technologies, and AI communication networks based on 5G and 6G.


This move can be seen as NVIDIA’s effort to expand its leadership in the AI ecosystem. The strategy is interpreted as NVIDIA seeking to maintain its strong presence in the AI infrastructure market even if some companies opt for custom processors from other vendors.


Jacob Born, a researcher at eMarketer, told Reuters, “NVIDIA will now have access to Marvell’s semi-custom silicon and advanced optical interconnect capabilities,” adding, “This will help scale data center-level AI systems, where bandwidth and power efficiency are critical bottlenecks.”


He also emphasized, “This collaboration will broaden the NVIDIA ecosystem to include more specialized silicon,” and, “It will help NVIDIA maintain a critical entry point amid increasingly diverse AI workloads.”


This investment is also an example of NVIDIA opening up its ecosystem to certain competitors. In May of last year, NVIDIA CEO Jensen Huang announced the company would open access to its AI servers, allowing various partners to deploy customer-specific chips in data centers.


Matt Murphy, Chairman and CEO of Marvell, stated, “Our collaboration with NVIDIA underscores the growing importance of high-speed connectivity, optical interconnects, and accelerated infrastructure in the expansion of AI infrastructure. By connecting Marvell’s strengths to NVIDIA’s AI ecosystem, we are enabling customers to build more efficient AI infrastructure.”



CEO Jensen Huang also stressed, “The world is rapidly accelerating the development of AI factories,” adding, “Together with Marvell, we are enabling customers to build and scale specialized AI computing by leveraging NVIDIA’s AI infrastructure ecosystem.”


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing