Cisco Unveils New Silicon One G200-Powered Switches for AI/ML Infrastructure

Cisco Introduces Silicon One G200-Powered Switches for AI/ML Networks

Cisco has unveiled new high-performance switches powered by its Silicon One G200 chip, designed to meet the growing demands of artificial intelligence and machine learning (AI/ML) workloads in enterprise and hyperscale data centers. The announcement comes as the AI industry experiences rapid growth and increasing network infrastructure requirements 1

1

Key Features of the New Switches

The newly introduced switches include the Cisco 8122-64EH/EHF and the Cisco Nexus 9364E-SG2, both powered by the Cisco Silicon One G200 chip. These switches offer:

1. High-density 800G connectivity with 64 ports in a 2RU form factor
2. Support for various port speeds including 400G, 200G, and 100G
3. Advanced load balancing and fault detection capabilities
4. Compatibility with open network operating systems like SONiC
5. Programmability and extensive API support for efficient network management 2

   2

Meeting AI/ML Network Demands

The new switches are designed to address the specific needs of AI/ML networks, which require:

1. Scalable and programmable high-radix switches
2. Low-power consumption
3. Advanced load balancing and observability
4. Support for flatter network topologies to reduce layers and improve performance 3

   3

Industry Adoption and Partnerships

Meta has announced plans to deploy the OCP-inspired Cisco 8501, which combines the Cisco Silicon One G200 with Cisco-designed hardware. This adoption by a major tech company underscores the potential of Cisco's new offerings in the AI/ML infrastructure space 1

1

Related Stories

Cisco unveils Silicon One G300 chip to challenge Broadcom and Nvidia in AI networking race

Cisco Unveils Silicon One P200 Chip to Connect AI Data Centers Across Vast Distances

Cisco Unveils New AI-Ready Data Center Infrastructure Solutions

Cisco and NVIDIA Expand AI Factory with Rack-Scale Computing to Meet Surging AI Data Center Demand

Benefits for Cloud Providers and Enterprises

The new switches offer several advantages for cloud providers and enterprises:

1. Improved cost control and reduced complexity
2. Lower latency and power consumption
3. Enhanced scalability for AI/ML workloads
4. Flexibility to meet critical cloud network infrastructure requirements
5. Support for both front-end and back-end network deployments 2

   2

Future-Proofing with UEC Compliance

Cisco's new switches are Ultra Ethernet Consortium (UEC) ready, complying with fabric baseline requirements such as PFC, ECN, and multiple traffic classes. This ensures robust performance for AI Ethernet networks and allows for adaptation to evolving UEC standards 3

3

As the AI industry continues to evolve rapidly, Cisco's latest offerings demonstrate the company's commitment to providing cutting-edge network infrastructure solutions that can meet the demands of next-generation AI/ML workloads while offering flexibility and scalability for enterprises and cloud providers alike.