As enterprises increasingly adopt AI, securing AI factories—where complex, agentic workflows run—has become critically important. NVIDIA addresses this need with its new NVIDIA DOCA software framework, part of the NVIDIA cybersecurity AI platform, which brings runtime cybersecurity to every AI factory. Running on the NVIDIA BlueField networking platform, DOCA Argus operates on each node to promptly detect and respond to threats targeting AI workloads, seamlessly integrating with enterprise security systems to provide instant threat insights. DOCA Argus enables real-time runtime threat detection through advanced memory forensics, achieving detection speeds up to 1, 000 times faster than existing agentless solutions without affecting system performance. Unlike traditional tools, Argus runs independently of the host, requiring no agents or host-based resource integration. This agentless, zero-overhead design boosts system efficiency and ensures resilient security in diverse AI compute environments, including containerized and multi-tenant infrastructures. Operating outside the host makes Argus invisible to attackers, even if the system is compromised. Cybersecurity teams can easily integrate DOCA Argus with SIEM, SOAR, and XDR platforms, enabling continuous monitoring, automated threat mitigation, and enhancing existing cybersecurity capabilities for AI infrastructure. The NVIDIA BlueField platform is a foundational security component for AI factories, delivering built-in, data-centric protection at scale. By combining BlueField’s acceleration with DOCA Argus’ proactive detection, enterprises secure AI factories without compromising performance or efficiency. Cisco partners with NVIDIA to offer a Secure AI Factory built on NVIDIA architecture, simplifying enterprise AI infrastructure deployment and protection at scale.

Their integrated security approach embeds runtime protection at every layer of the AI factory from the outset, rather than adding it post-deployment. “Now is the time for enterprises to advance AI, but the key to unlocking innovation and broad adoption is safety and security, ” said Jeetu Patel, Cisco’s EVP and chief product officer. “NVIDIA and Cisco provide enterprises with the infrastructure to confidently scale AI while protecting their most valuable data. ” DOCA Argus and BlueField are integral to the NVIDIA cybersecurity AI platform—a comprehensive, accelerated computing platform designed for AI-driven protection. It merges BlueField’s data-centric security and Argus’ real-time threat detection with NVIDIA AI Enterprise software—including the Morpheus cybersecurity AI framework—to deliver comprehensive visibility and control across AI factories. The platform leverages agentic AI to autonomously perceive, reason, and respond to threats in real time. With enterprises overwhelmed by massive data volumes, pinpointing genuine threats is challenging. The rise of agentic AI—AI models and autonomous agents connecting data, applications, and users at scale—offers novel insights but requires advanced protection able to keep pace. DOCA Argus is optimized using insights from NVIDIA’s security team, focusing on validated threats and known threat actors to minimize false positives. This approach provides actionable intelligence that reduces alert fatigue and streamlines security operations. Specifically designed to secure containerized workloads like NVIDIA NIM microservices, Argus incorporates real-world threat intelligence to protect all layers of the AI application stack. “Cyber defenders need robust tools to protect AI factories, the foundation for agentic reasoning, ” said David Reber, NVIDIA’s chief security officer. “The DOCA Argus framework delivers real-time security insights for autonomous detection and response, empowering defenders with actionable intelligence and a data advantage. ”