Marvell and NVIDIA Collaborate to Enhance AI Infrastructure with Custom Silicon Solutions

Marvell and NVIDIA Join Forces for Advanced AI Infrastructure

Marvell Technology, Inc. (NASDAQ: MRVL) has announced a strategic collaboration with NVIDIA Corp (NASDAQ: NVDA) to offer NVIDIA's NVLink Fusion technology to customers using Marvell's custom cloud platform silicon 1

1

. This partnership aims to revolutionize AI infrastructure by providing greater flexibility and accelerating deployment times for hyperscalers.

NVLink Fusion: Bridging Custom Silicon and AI Scalability

NVLink Fusion is NVIDIA's innovative offering that integrates custom XPU (any type of processor) silicon with NVIDIA NVLink connectivity, rack-scale hardware architecture, and software 2

2

. This technology allows hyperscalers to seamlessly deploy customized capabilities across their AI data center infrastructure while leveraging existing investments in NVLink architecture.

Key Benefits for Hyperscalers

The collaboration between Marvell and NVIDIA offers several advantages for cloud providers:

1. Scalability: NVLink enables cloud providers to scale out AI factories to millions of custom XPUs 1

   1

   .
2. Integration: Proprietary XPUs can be integrated into NVIDIA's rack-scale systems and end-to-end networking platform 2

   2

   .
3. Performance: The core technology of NVLink Fusion delivers up to 1.8TB/s of bidirectional bandwidth 2

   2

   .
4. Flexibility: Customers gain greater choice in developing next-generation AI infrastructure 2

   2

   .

Marvell's Custom Platform Strategy

Marvell's approach to custom silicon solutions involves:

1. Unique semiconductor designs
2. Innovative approaches
3. Expertise in system and semiconductor design
4. Advanced process manufacturing
5. Comprehensive portfolio of semiconductor platform solutions and IP 2

   2

This strategy allows Marvell to create platforms in collaboration with customers that transform infrastructure performance, efficiency, and value.

Industry Perspectives

Nick Kucharewski, Senior Vice President and General Manager of Marvell's Cloud Platform Business Unit, emphasized the collaboration's goal: "Marvell and NVIDIA are working together to advance AI factory integration. 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" 2

2

.

Shar Narasimhan, Director of Accelerated Computing at NVIDIA, highlighted the changing landscape: "The computing landscape is being reshaped as AI is no longer an application—it is foundational to modern data centers. NVLink Fusion extends NVIDIA's open platform to partners like Marvell, enabling hyperscalers to scale out AI factories to millions of GPUs" 2

2

.

Market Impact and Future Outlook

Despite the positive news, Marvell's stock has faced challenges, plunging 44% year-to-date due to the impact of Trump administration tariff policies on the semiconductor industry 1

1

. The company has postponed its Investor Day and updates on its AI strategy from June 2025 to an unspecified date in 2026, citing ongoing macroeconomic uncertainty 1

1

.

However, industry analysts remain optimistic. CNBC's Jim Cramer has recommended buying Marvell stock following the company's announcement of successful interoperability between its Structera portfolio of Compute Express Link (CXL) devices and AMD EPYC CPUs and 5th Gen Intel Xeon platforms 1

1

.