Simplify AI to meet your business needs
Trusted, high-performance AI from a cloud-managed, full-stack AI infrastructure that is compliant with NVIDIA Enterprise Reference Architectures (RA)
Cisco Nexus Hyperfabric offerings for AI
BYO AI with Nexus Hyperfabric
Empower your AI innovation with a cloud-managed network fabric that seamlessly works with your existing compute, GPU, and storage resources.
Key components:
- Cisco Nexus Hyperfabric cloud-managed controller
- Cisco 6000 Series Switches with 800G Ethernet and Cisco Silicon One
- Network fabric managed via cloud controller
Nexus Hyperfabric AI
Accelerate AI innovation with a cloud-managed, on-premises, full-stack solution that simplifies deployment and management while delivering a high-performance AI infrastructure.
Key components:
- Cisco Nexus Hyperfabric cloud-managed controller
- Cisco 6000 Series Switches with Silicon One technology for AI workloads
- 800 GbE backend fabric with low-latency, lossless networking
- UCS C885A M8 Rack Server and NVIDIA Tensor Core GPUs, starting with NVIDIA H200 SXM with NVLink
- NVIDIA BlueField-3 DPUs and BlueField-3 SuperNIC
- Integration with optional VAST Data Platform for scalable, multi-tenant, zero-trust, all-flash storage
- Cloud-delivered software upgrades with user-driven update schedules
- Assertion-based monitoring, telemetry, and real-time health checks spanning network to compute NIC
FAQs for Cisco Nexus Hyperfabric AI
Cisco Nexus Hyperfabric AI is a cloud-managed full stack AI infrastructure solution delivered as hardware + software + service. Using a cloud controller managed by Cisco, customers easily design, deploy, and manage their network fabric, GPU servers, and storage servers. It reinvents the IT operations lifecycle of the AI infrastructure in the data center by simplifying every step of the process and ensures repeatable and predictable outcomes by IT generalist, application, and DevOps teams. The vertical stack solution consists of purpose-built hardware, software, cloud management, Day-2 automation, and Cisco support. Cisco Nexus Hyperfabric AI is best suited for customers looking to build out their private cloud AI infrastructure.
Cisco Nexus Hyperfabric AI is a complete, turnkey solution designed for ease of deployment and accelerated day-2 value. It is standards-based and will interoperate with existing data-center fabrics connected, for example, through border gateways. As a turnkey solution, Cisco Nexus Hyperfabric AI will run only on specific Cisco hardware, and only that hardware can be part of a Nexus Hyperfabric AI cluster.
Cisco Nexus Hyperfabric AI consists of five primary components:
- Cloud Controller: a scalable, globally distributed multitenant cloud service that is used to design, plan, control, upgrade, and monitor fabrics using a browser or APIs known as Cisco Nexus Hyperfabric.
- Cisco 6000 Series Switches: installed with Cisco Nexus Hyperfabric AI–managed software, they connect to the cloud for centralized real-time visibility and control.
- Cisco UCS® NVIDIA GPU/DPU Servers: Cisco UCS-C885A-M8-CN1 server that packs 8 NVIDIA H200 GPUs, with BlueField-3 DPU (3240H) and SuperNIC (3140H) connected using Cisco® optics that can efficiently run training/inferencing/fine-tuning jobs.
- UCS-VAST Storage (Optional): Cisco UCS-C225-M8N-1P servers pack 8 drives with each of the servers with a cluster consisting of 11 servers that can be easily expanded. The specs start with 1PB of storage that can be used for training, fine-tuning, inferencing, Retrieval-Augmented Generation (RAG), and other data engineering work.
- NVAIE: Integration and access to the NVIDIA AI for Enterprise (NVAIE) software stack from get-go allows access to different model training software and data catalogues so that engineers and scientists can get started immediately.