Marvell Technology announced a collaboration with NVIDIA to provide NVIDIA NVLink Fusion technology to customers using Marvell custom cloud platform silicon. NVLink Fusion is a revolutionary new offering from NVIDIA that combines proprietary XPU silicon with NVIDIA NVLink connectivity, rack-scale hardware architecture, software, and other technology, providing customers more flexibility and choice when developing next-generation AI infrastructure.
Marvell’s custom platform strategy aims to achieve breakthrough achievements through unique semiconductor designs and novel methodologies.
By combining expertise in system and semiconductor design, advanced process manufacturing, and a comprehensive portfolio of semiconductor platform solutions and IP—including electrical and optical serializer/deserializers (SerDes), die-to-die interconnects for 2D and 3D devices, advanced packaging, silicon photonics, co-packaged copper, custom high-bandwidth memory (HBM), system-on-chip (SoC) fabrics, optical IO, and compute fabric interfaces such as PCIe Gen 7—Marvell. Marvell is able to create platforms in collaboration with customers that transform infrastructure performance, efficiency and value.
Marvell custom silicon with NVLink Fusion provides hyperscalers with a streamlined path to custom scale-up solutions that fulfill the stringent requirements of model training and agentic AI inference—where outputs are guided by learned knowledge and reasoning.
Hyperscalers may now seamlessly deploy these tailored capabilities throughout their AI data center infrastructure while leveraging their NVLink architectural investments.
NVLink enables cloud providers to easily scale out AI factories to millions of custom XPUs by efficiently integrating their proprietary XPUs into NVIDIA’s rack-scale systems and the NVIDIA end-to-end networking platform. The basic technology of NVLink Fusion is a chiplet that can supply up to 1.8TB/s of bidirectional bandwidth.
Leadership Comments
“Marvell and NVIDIA are working together to advance AI factory integration,” said Nick Kucharewski, senior vice president and general manager, Cloud Platform Business Unit at Marvell. “Through this collaboration, we offer customers the flexibility to rapidly deploy scalable AI infrastructure with the bandwidth, performance and reliability required to support advanced AI models.”
“The computing landscape is being reshaped as AI is no longer an application—it is foundational to modern data centers,” said Shar Narasimhan, Director of Accelerated Computing at NVIDIA. “NVLink Fusion extends NVIDIA’s open platform to partners like Marvell, enabling hyperscalers to scale out AI factories to millions of GPUs, using custom silicon, NVIDIA’s rack-scale systems and the NVIDIA end-to-end networking platform, to meet the world’s accelerating demand for intelligence.”
