NVIDIA 920-9N42C-00RB-7C0 Spectrum-4 Data Center Switch

• AI and machine learning cluster networking
• Hyperscale cloud data center fabrics
• Spine-leaf Ethernet architecture deployments
• High-performance computing (HPC) environments
• Large enterprise core network aggregation
• Multi-tenant cloud infrastructure platforms
• Storage networking and data replication
• East-west traffic optimization in modern data centers
• High-density virtualization environments
• Service provider and telecom backbone networks

• 64-Port Native 400GbE High-Density Architecture
• 25.6Tbps Non-Blocking Switching Performance
• 33.3 Billion Packets Per Second Forwarding
• Flexible Breakout Support for Multi-Speed Networks
• Enterprise-Class Redundant Power Protection
• Hot-Swappable Fan Modules for High Availability
• Integrated x86 Management Processing Platform
• Large 160MB Shared Packet Buffer Design
• NVMe SSD Storage for Enhanced Operations
• Optimized Airflow for Modern Data Centers

• Brand: NVIDIA 
• Product Family: Spectrum-4 Ethernet Switch
• Model Number: 920-9N42C-00RB-7C0
• Switch Type: Data Center Ethernet Switch
• Architecture: Spectrum-4 ASIC
• Form Factor: Rack-Mount Network Switch
• Primary Connectivity: Ethernet
• Connector Configuration: 64 × QSFP56-DD 400GbE, 2 × SFP28 25GbE
• Switching Capacity: 25.6 Tbps
• Forwarding Performance: 33.3 Billion Packets Per Second (Bpps)
• Packet Buffer Size: 160 MB
• CPU Type: Hexa-Core x86 Processor
• System Memory: 32 GB
• SSD Storage Capacity: 160 GB
• Management Port: 1 × RJ45 100/1000Mb/s
• Console Port: RJ45 Serial Console
• USB Interface: USB Type 3
• Power Supply Configuration: Dual Hot-Swappable AC Power Supplies
• Power Redundancy: 1+1 Redundant
• Number of Fan Modules: 4
• Fan Redundancy: N+1 Redundant
• Cooling Design: Hot-Swappable Fans
• Airflow Direction: Reverse Airflow

WHY CHOOSE THIS PRODUCT?

• Delivers 25.6Tbps switching capacity for bandwidth-intensive workloads.
• Supports up to 64 native 400GbE connections for dense network deployments.
• Enables flexible breakout configurations for multi-speed migration strategies.
• Redundant power and cooling reduce downtime risks in mission-critical environments.
• Optimized for AI, machine learning, cloud, and HPC infrastructures.
• High-density design helps lower rack space requirements and operational costs.
• Integrated x86 management resources simplify monitoring and administration.
• Future-ready architecture supports evolving data center networking demands.