Marvell and NVIDIA Collaborate on AI Infrastructure Solutions

May 19, 2025
Marvell Technology and NVIDIA have announced a collaboration to offer custom solutions for advanced AI infrastructure, utilizing NVIDIA's NVLink Fusion technology.

Marvell Technology and NVIDIA have announced a collaboration to provide custom solutions for advanced AI infrastructure, announced in a press release. The partnership will leverage NVIDIA's NVLink Fusion technology, which integrates custom XPU silicon with NVIDIA's connectivity and hardware architecture, to offer greater flexibility and choice for developing next-generation AI infrastructure.

The collaboration aims to accelerate the deployment of scalable AI infrastructure by combining Marvell's expertise in semiconductor design with NVIDIA's advanced networking platform. This integration allows hyperscalers to deploy custom scale-up solutions that meet the demands of AI model training and inference, while also utilizing existing architecture investments in NVLink.

NVLink Fusion, a key component of this collaboration, provides a chiplet delivering up to 1.8TB/s of bidirectional bandwidth, enabling cloud providers to scale out AI factories to millions of custom XPUs. This partnership marks a significant step in advancing AI factory integration, offering customers the bandwidth, performance, and reliability needed to support advanced AI models.

We hope you enjoyed this article

Consider subscribing to one of our newsletters like Silicon Brief or Daily AI Brief.

Also, consider following us on social media:

AI Chips & Datacenters AI Brief AI Brief (X)
Free newsletter

Silicon Brief

Weekly coverage of AI hardware developments including chips, GPUs, cloud platforms, and data center technology.

AI Token Webinar

Whitepaper

Tensordyne Napier: What If One Rack Could Do the Work of Nine?

Tensordyne

This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.

Read more

You may also like

Free, six days a week

Daily AI Brief: the AI news that matters, in your inbox.