Memory has become the defining constraint of modern AI infrastructure. Large language models, in-memory databases, and deep learning recommendation models all share the same bottleneck: there is never enough DRAM. Compute Express Link (CXL) was designed to break that bottleneck but a CXL memory expander is only as valuable as the breadth of platforms it can run on.

That is why ecosystem enablement is a core pillar of Marvell's CXL strategy. Today we are marking a new milestone: successful interoperability of the Marvell® Structera™ X CXL memory-expansion controller with the NVIDIA Vera CPU platform.

The Structera CXL Portfolio: Built for the Full Stack

Marvell is the only provider with a complete end-to-end CXL portfolio spanning memory expansion (Structera X), near-memory acceleration (Structera A), and rack-scale memory pooling (Structera S CXL switch), giving customers a single, integrated architecture from controller to fabric.

Together, Structera X, A, and S form a complete CXL fabric architecture, from per-server memory expansion to rack-scale pooling, that no competitor can match end-to-end.

Interoperability: The Foundation of Deployment at Scale

A portfolio is only as strong as its ecosystem. Marvell has invested systematically in validating Structera across every major CPU architecture and memory supplier.

In April 2025, Marvell announced successful interoperability of the Structera CXL portfolio with AMD EPYC™ CPUs and 5th Gen Intel® Xeon® Scalable platforms—the result of extensive joint testing across configurations, workloads, and operating conditions. That made Structera the only CXL 2.0 product family with completed interoperability across both leading x86 CPU architectures.

In September 2025, Marvell extended that leadership to memory, completing interoperability testing with DDR4 and DDR5 solutions from Micron Technology, Samsung Electronics, and SK Hynix. Structera became the only CXL 2.0 product family validated across both leading CPU architectures and all three major memory suppliers.

Now Expanding NVIDIA Vera

The data center CPU landscape is no longer exclusively x86. New server platforms are gaining significant traction across cloud-native workloads, AI inference, storage, and high-performance computing. The collaboration with NVIDIA led to the interoperability testing of the Structera X CXL Type 3 memory expander on the NVIDIA Vera platform.

The test covered the full range of CXL 2.0 compliance and feature validation: device enumeration, CXL and FLR resets, Data Object Exchange (DOE) and CDAT delivery, and the complete CXL mailbox command sets including event management, health monitoring, and firmware update. The results confirm that Structera X delivers consistent CXL 2.0 compliance and feature support across multiple platforms—a critical requirement for customers deploying heterogeneous server fleets.

What This Means for Customers

Cloud operators, server and storage OEMs, and enterprise customers deploy diverse CPU infrastructure. They need CXL solutions that work seamlessly across all of them, without re-qualification for each new platform. The interoperability-first approach of Marvell directly reduces that friction.

With validated support across Intel Xeon, AMD EPYC, and now NVIDIA Vera, Structera is the most broadly validated CXL solution in the industry. Combined with memory interoperability across Micron, Samsung, and SK hynix, and a complete portfolio spanning expansion, acceleration, and pooling, Marvell gives customers the flexibility to deploy CXL across diverse hardware configurations—today and as their infrastructure evolves.

To learn more about Marvell Structera CXL solutions, visit https://www.marvell.com/products/cxl.

# # #

This blog contains forward-looking statements within the meaning of the federal securities laws that involve risks and uncertainties. Forward-looking statements speak only as of the date they are made. Marvell, the M logo, and Structera are trademarks of Marvell or its affiliates. CXL and Compute Express Link are registered trademarks of the CXL Consortium. AMD and EPYC are trademarks of Advanced Micro Devices, Inc. Intel and Xeon are trademarks of Intel Corporation or its subsidiaries.