NVIDIA and Cisco Unveil Nexus HyperFabric for Generative AI Infrastructure

NVIDIA and Cisco have announced a new collaborative effort to enhance enterprise generative AI infrastructure with the launch of the Nexus HyperFabric AI cluster solution. This development aims to provide enterprises with a robust pathway to operationalize generative AI, according to NVIDIA Blog.

Empowering Generative AI with HyperFabric

The Nexus HyperFabric is designed to handle the extensive data processing, computational power, and networking bandwidth required by generative AI models. The solution integrates NVIDIA's accelerated computing technologies and AI software with Cisco's AI-native networking capabilities and the VAST Data Platform.

Kevin Wollenweber, senior vice president and general manager of data center and provider connectivity at Cisco, highlighted the importance of this collaboration by stating, “Enterprise applications are transforming into generative AI applications, significantly increasing data processing requirements and overall infrastructure complexity. Together, Cisco and NVIDIA are advancing HyperFabric to advance generative AI for the world’s enterprises so they can use their data and domain expertise to transform productivity and insight.”

Powering a Full-Stack AI Fabric

At the core of the Nexus HyperFabric solution are NVIDIA Tensor Core GPUs, which are essential for processing large datasets. The solution also includes NVIDIA AI Enterprise, a cloud-native software platform that streamlines the development and deployment of production-grade AI applications. This platform ensures optimized performance, security, and API stability for enterprise AI deployments.

Additionally, NVIDIA NIM inference microservices are part of the package, facilitating the deployment of foundational models while maintaining data security. These microservices bridge the gap between complex AI development and operational needs, supporting the entire AI journey from ideation to production-scale deployment.

The Cisco Nexus HyperFabric AI cluster also integrates NVIDIA BlueField-3 SuperNICs and DPUs, which enhance system performance and security. The SuperNICs provide advanced network capabilities, ensuring seamless, high-speed connectivity. BlueField-3 DPUs offload, accelerate, and isolate infrastructure services, resulting in a more efficient AI solution.

AI-Powered Security with Cisco Hypershield

BlueField-3 DPUs can also run security services such as Cisco Hypershield, an AI-native, hyperdistributed security architecture. Hypershield shifts security measures closer to the workloads needing protection, illustrating another area of innovation between Cisco and NVIDIA focused on AI-powered security solutions.

Showcasing at Cisco Live

The capabilities of the Nexus HyperFabric AI cluster will be showcased at Cisco Live in Las Vegas, running through June 6. Attendees can visit the Cisco AI Hub to see NVIDIA AI technologies in action and learn best practices for enterprise AI deployment.

Key sessions include:

• Keynote Deep Dive: “Harness a Bold New Era: Transform Data Center and Service Provider Connectivity” with NVIDIA and Cisco executives — June 5, 1-2 p.m. PT
• AI Hub Theater Presentation: “Accelerate, Deploy Generative AI Anywhere With NVIDIA Inference Microservices” — June 4, 2:15-2:45 p.m. PT
• WWT AI Hub Booth: Thought leadership interview with NVIDIA and WWT executives — June 5, 10-11 a.m. PT
• NetApp Theater: “Accelerating Gen AI With NVIDIA Inference Microservices on FlexPod” — June 5, 1:30-1:40 p.m. PT
• Pure Storage Theater: “Accelerating Gen AI With NVIDIA Inference Microservices on FlashStack” — June 5, 2-2:10 p.m. PT