Red Hat & NVIDIA boost AI security with BlueField integration
Red Hat has announced support for Red Hat OpenShift on NVIDIA BlueField data processing units, aiming to help organisations deploy and run artificial intelligence workloads with greater security and efficiency.
This support is designed to address growing pressures on business infrastructure resulting from the rise in AI adoption. As AI applications become more prevalent, they increasingly compete with vital infrastructure services for limited resources, which can lead to potential performance bottlenecks and heightened security risks.
The collaboration between Red Hat and NVIDIA will provide a unified, cloud-native platform, separating AI applications from critical infrastructure tasks. This design aims to deliver improved scalability, security, and performance for data centres managing next-generation AI environments.
Technical features
Red Hat OpenShift on NVIDIA BlueField offers several enhancements. It provides optimised resource utilisation by offloading networking services from the central processing unit to the data processing unit. This frees up CPU resources for dedicated AI workloads, supporting more efficient use of computational power.
Acceleration of data plane and storage traffic is enabled by shifting cryptography and storage processing to the DPU. This is supported by technologies such as NVMe over Fabrics (NVMe-oF) and an accelerated Open vSwitch (OVS) data path, which together improve performance for high-demand AI workloads.
Advanced cloud networking capabilities are also included. Red Hat OpenShift, with distributed routing powered by BlueField, aims to improve secure multi-tenancy, network scalability, and traffic efficiency across multiple clusters. This approach is designed to simplify and enhance network operations in large data centre environments.
Red Hat reports that isolating infrastructure workloads on BlueField, including networking and security tasks, can reduce the attack surface for applications. By offloading infrastructure tasks to the BlueField DPU, the company claims that this approach contributes to a better overall security posture for organisations relying on AI-driven applications.
Ongoing collaboration
Support for NVIDIA BlueField on Red Hat OpenShift will initially be available as a technical preview. Red Hat has indicated that it will continue to collaborate with NVIDIA, aiming to integrate further support for the NVIDIA DOCA software framework and compatible third-party network functions for use with Red Hat OpenShift on BlueField.
Future developments are planned, with the upcoming NVIDIA BlueField-4 expected to deliver increased acceleration, deeper DOCA integration, and greater performance for cloud-native AI factories. Red Hat is also working with NVIDIA to support Spectrum-X Ethernet networking to enable high-performance connectivity for AI workloads across distributed cloud environments.
Industry perspectives
Ryan King, vice president, AI and Infrastructure, Partner Ecosystem Success, Red Hat, commented, "As the adoption of generative and agentic AI grows, the demand for advanced security and performance in data centres has never been higher, particularly with the proliferation of AI workloads. Our collaboration with NVIDIA to enable Red Hat OpenShift support for NVIDIA BlueField DPUs provides customers with a more reliable, secure and high-performance platform to address this challenge and maximise their hardware investment."
Justin Boitano, vice president, Enterprise Products, NVIDIA, said, "Data-intensive AI reasoning workloads demand a new era of secure and efficient infrastructure. The Red Hat OpenShift integration of NVIDIA BlueField builds on our longstanding work to empower organisations to achieve unprecedented scale and performance across their AI infrastructure."
The collaboration between Red Hat and NVIDIA follows ongoing efforts in the technology sector to adapt infrastructure to support the growing intensity and complexity of AI applications. By leveraging DPUs, organisations are expected to benefit from enhanced resource allocation, improved security, and accelerated performance for mission-critical workloads.
