At the Future of Memory and Storage (FMS) trade show, Marvell won the AI Infrastructure award for its Photonic Fabric™ technology, a transformational optical interconnect platform that overcomes the bandwidth and memory bottlenecks limiting AI infrastructure. By replacing electrical interconnects with optical I/O across package-, server- and rack-scale architectures, Marvell® Photonic Fabric™ enables scalable AI systems with dramatically higher bandwidth, lower latency, reduced power consumption and unprecedented memory scalability.

Memory bandwidth, memory locality and data movement are pressing constraints on AI training and inference scalability and token efficiency. Traditional architectures were not designed to support hundreds or thousands of accelerators across multiple racks. Photonic Fabric technology facilitates a new generation of disaggregated memory architectures where memory capacity is no longer constrained by the physical boundaries of individual servers or packages.

The Photonic Fabric technology platform replaces traditional electrical connectivity with a unified optical interconnect system tailored for scale-up networks, establishing a new architectural foundation where compute, memory, and storage scale together.

As hyperscalers transition toward AI factories containing hundreds of thousands of interconnected accelerators and associated memory systems, Photonic Fabric technology provides the architectural breakthrough required to scale AI efficiently, sustainably, and economically.

Technology Highlights

• Enables significantly more memory bandwidth and capacity while reducing latency and power consumption
• Enables disaggregated memory and compute architectures for next-generation AI factories
• Replaces electrical interconnects with optical I/O across package-, server-, and rack-scale AI infrastructure
• Eliminates die-edge I/O bottlenecks, enabling greater compute and HBM density
• Establishes the optical foundation for scalable, energy-efficient AI infrastructure

Photonic Fabric technology fundamentally redefines how AI factories are architected, and is part of the company’s industry-leading AI silicon connectivity portfolio including scale-up, out-, and across optical technologies, silicon photonics, and switch silicon. By transforming data movement from AI's greatest bottleneck into a scalable resource, Marvell is establishing the foundation for the next generation of AI infrastructure.

# # #

This blog contains forward-looking statements within the meaning of the federal securities laws that involve risks and uncertainties. Forward-looking statements include, without limitation, any statement that may predict, forecast, indicate or imply future events or achievements. Actual events or results may differ materially from those contemplated in this blog. Forward-looking statements are only predictions and are subject to risks, uncertainties and assumptions that are difficult to predict, including those described in the “Risk Factors” section of our Annual Reports on Form 10-K, Quarterly Reports on Form 10-Q and other documents filed by us from time to time with the SEC. Forward-looking statements speak only as of the date they are made. Readers are cautioned not to put undue reliance on forward-looking statements, and no person assumes any obligation to update or revise any such forward-looking statements, whether as a result of new information, future events or otherwise.