Marvell, Microsoft, and Oracle Collaborate on Standardized Data Center Hardware Security
Marvell, Microsoft, and Oracle are partnering through the Open Compute Project to develop an industry standard for integrated Hardware Security Modules (HSMs) in data centers.
Marvell Technology, Microsoft, and Oracle have announced a significant collaboration aimed at standardizing hardware security within data center infrastructure. The initiative, aligned with the Open Compute Project (OCP), focuses on creating an integrated Hardware Security Module (HSM) that meets the stringent NIST FIPS 140-3 Level 3 compliance standard. This effort seeks to bolster security for cloud and hyperscale environments by establishing common, robust hardware security specifications.
To drive this standardization, the three tech giants have established a new OCP working group under the project's Security Track. The group's primary objective is to define open specifications for an integrated HSM. This module will be designed for secure key caching and comprehensive cryptographic key lifecycle management directly at the server level. Such an approach promises to reduce latency, enhance scalability, and significantly strengthen the security posture of cloud computing workloads.
Currently, many cloud and enterprise servers rely on either network-attached HSMs or software-only key management solutions. These methods often expose sensitive cryptographic keys by requiring them to traverse the network or reside in unprotected memory, making them vulnerable to security attacks. The new OCP working group aims to address this by defining an open specification for embedding a hardware security boundary, a secure key vault, and an attestation chain directly into the server itself.
This embedded approach ensures that cryptographic operations are performed where the data is processed, cryptographic keys remain within a certified hardware boundary, and workload integrity can be cryptographically proven to any relying party. "An integrated HSM will allow cryptographic key operations to execute where the workload runs, satisfying both regulatory compliance and zero-trust architecture requirements, all without compromising higher performance for security," stated Cary Ussery, senior vice president and general manager, Multimarket Business Group, at Marvell.
Mark Russinovich, CTO, deputy CISO and technical fellow at Microsoft Azure, emphasized the critical timing of this initiative. "As AI transforms cybersecurity and quantum technology reshapes cryptography, protecting the keys and credentials at the heart of our digital infrastructure has never been more important," he said. "Through the Open Compute Project, industry leaders are coming together around a fully open-source integrated HSM, combining our collective innovation to advance interoperability, accelerate the adoption of post-quantum cryptography, and shape the future of hardware security."
Oracle echoed this sentiment, with Pradeep Vincent, executive vice president, Compute and AI Infrastructure, noting, "An open specification for integrated Hardware Security Module (HSM) capabilities is an important step toward enabling cryptographic operations where data is processed, protecting keys within a hardware security boundary and supporting cryptographic verification of workload integrity." He added that the effort aims to establish a common foundation for industry-wide adoption.
As founding members, Marvell, Microsoft, and Oracle will collaborate with the broader OCP community to develop and publish a joint technical specification. An initial draft is anticipated around the 2026 OCP Global Summit, with version 1.0 to follow. The companies plan to showcase this new technology at their respective booths and during Security Track presentations at the upcoming OCP Global Summit in San Jose, California.
This collaboration represents a significant step towards a more secure and standardized hardware foundation for the rapidly evolving cloud and hyperscale data center landscape, particularly as it grapples with the increasing demands of AI and the future threat landscape posed by quantum computing.