en.Wedoany.com Reported - DDN has updated its AI data intelligence platform to accelerate the deployment of agentic AI.

This update aims to strengthen data governance and security, reduce operational complexity, and maximize GPU efficiency in enterprise-grade AI infrastructure. New features include real-time observability, policy-based controls, secure multi-tenant isolation, and AI-native data orchestration.

These features are designed for large-scale training, inference, and autonomous AI workloads, helping companies transition AI projects from pilot to production while improving overall performance and return on investment.

DDN's platform update aligns with NVIDIA's product roadmap, including the BlueField-4 STX architecture announced at GTC earlier this year. DDN has adopted this reference architecture, which also relies on DOCA software stack components (such as the recently launched Argus) to help enterprise customers scale secure AI environments for training and inference. According to the vendor, the combined solution provides inline security, memory observability, and policy-based protection for AI-native storage and agentic AI workloads running at scale.

Jason Hardy, Vice President of Storage Technology at NVIDIA, stated that as enterprises move autonomous AI from pilot to production, a new generation of secure, high-performance data infrastructure is critical for managing the large-scale real-time demands of agentic workloads. Integrating Vera BlueField-4 STX and the DOCA security framework with DDN's AI-native data intelligence platform enables enterprises to operate secure, scalable AI factories for large-scale training and inference.

DDN already provides data infrastructure for large-scale AI environments, sovereign AI deployments, hyperscale cloud providers, and enterprise systems worldwide. Leveraging NVIDIA accelerated computing, the DDN platform aims to help enterprises build secure AI environments by combining high-performance data orchestration, multi-tenant isolation, and real-time services optimized for training, inference, vector databases, retrieval-augmented generation (RAG) pipelines, and autonomous workflows.

This release highlights an industry trend of implementing AI security at the infrastructure layer. This model ensures policy enforcement and protection occur directly within the AI data path, rather than relying solely on traditional host-based defenses. To capitalize on growing demand for AI infrastructure, DDN recently added capabilities to its Lustre platform, allowing users to share key-value (KV) caches to accelerate AI inference workloads. Announced at Google's annual Next event in April and co-managed by DDN and the hyperscaler, the solution uses a shared cache layer across inference clusters instead of retaining KV caches in each server's local memory. DDN and its hyperscaler partner claim this approach improves total inference throughput by up to 75%.

This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com