Marvell to Highlight Next-generation Accelerated Infrastructure at 2025 OCP Global Summit

Driving the Future of AI and Cloud Data Centers Through Open Ecosystem Innovation, Connectivity, Performance and Power Efficiency

SANTA CLARA, Calif. – October 7, 2025 – Marvell Technology, Inc. (NASDAQ: MRVL), a leader in data infrastructure semiconductor solutions, today announced it will highlight its latest advancements in accelerated infrastructure at the OCP Global Summit, taking place from October 13 to 16 at the San Jose Convention Center in San Jose, California.

As AI becomes the new center of gravity for data center architectures, the scale and complexity of these systems are pushing the limits of traditional design. To meet this challenge, Marvell is driving advancements in custom silicon, chiplet integration, next-generation memory architectures and tightly integrated connectivity. These advances support AI clusters at every scale, from AI servers to racks to entire campuses and multi-site topologies while improving time-to-deployment, lowering power consumption and reducing cost-per-bit.

During the 2025 OCP Global Summit, Marvell will highlight how its full-stack portfolio—spanning advanced silicon platforms, high-speed interconnects and network switches—is powering the next wave of AI infrastructure. Leveraging open collaboration, standards-based innovation and robust ecosystem partnerships, Marvell innovations are driving tomorrow’s complex AI and cloud data center deployments by maximizing performance, scalability, interoperability and energy efficiency.

Marvell experts will participate in multiple sessions at OCP 2025, sharing insights into how the company’s commitment to open, standards-based innovation is shaping the future of AI data centers. The Marvell program agenda can be found here.

In addition, Marvell will feature several technology demonstrations at the conference 
(Expo Hall, Booth B1), including:

• Co-packaged copper system with 2.5m direct attach cable (DAC)
• Co-packaged optics switch system
• Marvell® Structera™ CXL near-memory acceleration
• Marvell Structera CXL memory expansion and compression
• Marvell Alaska® P PCIe 6 retimer over cable for AI scale up
• 800G/1.6T active electrical cable (AEC) for AI scale up and scale out
• Marvell Ara 200G/lambda 1.6T PAM4 optical DSP for AI scale out
• Marvell Teralynx® switch telemetry API for AI infrastructure

关于 Marvell

为了通过数据基础设施技术实现全球互联的愿景，我们正在构建非常强大的基础解决方案： 我们与客户携手。 30 年来，Marvell 深受世界位居前列的技术公司的信赖，以客户当前需求和未来理想为设计路线，运用专业半导体解决方案实现对全球数据的移动、存储、处理与保护。 通过深度协作、践行透明化的沟通方式，我们致力于促进未来企业、云、汽车和运营商架构的不断转型与完善。

# # #

Marvell 和 M 徽标是Marvell或其附属公司的商标。请访问 www.marvell.com 网站获取Marvell商标的完整列表。其他名称与品牌可能是其他方财产。

本新闻稿包含联邦证券法界定的前瞻性声明，该等声明涉及风险和不确定性。 前瞻性声明包括但不限于任何可能预测、预估、间接提及或暗示未来事件、结果或成就的声明。 实际事件、结果或成就可能与本新闻稿的预期存在重大差异。 前瞻性声明仅为预测，且受难以预测的风险、不确定性和假设的影响，这些风险、不确定性和假设包括 Marvell 的 10-K 年度报表中的“风险因素”部分、10-Q 季度报表以及 Marvell 不时向美国证券交易委员会 (Securities and Exchange Commission, SEC) 提交的其他文件中提及的内容。前瞻性声明仅在发表当日有效。 谨提醒读者不要过分依赖前瞻性声明，无论是由于新信息、未来事件还是其他原因，任何人都没有义务更新或修改任何此类前瞻性声明。

For further information, contact:
George Millington
pr@marvell.com