Cisco 400G & 800G Data Center Switches
Deploy 400G and 800G switching powered by Cisco Silicon One — from fixed leaf switches to modular spine chassis for AI/ML, cloud, and enterprise data centers.
As AI/ML and cloud workloads drive unprecedented bandwidth demands, Cisco Nexus 400G and 800G switches deliver the performance, programmability, and scale your data center needs.
Cisco Silicon One architecture delivers industry-leading performance-per-watt, helping you meet net-zero targets without sacrificing throughput.
Reduce your footprint with higher port densities in fewer RUs. Scale from 12.8 Tbps to over 100 Tbps with a single, consistent operating model.
Purpose-built for RoCE, RDMA, and GPU-to-GPU traffic with adaptive routing, PFC/ECN, and ultra-low latency fabrics that keep your AI infrastructure moving.
Invest in a platform that scales from 400G fixed switches to 800G spine layers with a unified Silicon One architecture and consistent operating model.
The foundation of your high-density 400G data center fabric.
4RU Chassis
4RU modular chassis with 8 expansion slots
Supervisor
Supervisor module for chassis control plane
Switch Card
Internal fabric switch card for data movement
Choose the port speed and density that matches your data center workload.
400G
8-port 400G QSFP-DD expansion module
200G
16-port 200G QSFP56 expansion module
10/25/50G
22-port 10/25/50G SFP56 expansion module
High-density, 1RU and 2RU fixed-port switches for top-of-rack, leaf, and spine deployments.
400G
64x 400G QSFP-DD, 25.6 Tbps, Silicon One — ultra-high-density leaf or spine with MACsec on all ports.
Explore Nexus 9364D-GX2A
400G
32x 400G QSFP-DD, 25.6 Tbps — AI/ML optimized spine with ultra-low latency and advanced flow control for GPU fabrics.
Explore Nexus 9332D-GX2B
400G
32x 400G, deep on-chip shared buffer — first Nexus 9000 with deep buffer architecture for AI/ML and IP media.
Explore Nexus 9332D-H2R
400G
48x 400G QSFP-DD, 38.4 Tbps — line-rate MACsec on every port for secure AI/ML and financial trading fabrics.
Explore Nexus 9348D-GX2AThe next frontier in data center switching — 800G per port for AI/ML supercomputing, HPC, and hyperscale cloud.
800G
64x 800G OSFP ports, 51.2 Tbps — enterprise/HPC/telco spine designed for GPU cluster fabrics and AI supercomputing at scale.
Explore Nexus 9364E-SP2R
800G
Cisco’s first 800G enterprise data center switch. 64x 800G OSFP, power-optimized silicon, AI/ML lossless fabric with VXLAN BGP EVPN.
Explore Nexus 9364E-SG2One unified architecture from access to hyperscale core — delivering consistent forwarding, programmability, and security at every port speed.
One ASIC from access to hyperscale core. Consistent forwarding behavior, unified tooling, and a common programmability model.
P4-programmable data plane lets you customize forwarding behavior, add new protocols, and adapt to emerging workloads without hardware replacement.
Model-driven, real-time telemetry with sub-second granularity. Full visibility into traffic patterns, congestion events, and fabric health.
Line-rate 256-bit AES-GCM MACsec encryption on all ports — from 100G to 800G — keeps your fabric secure without any performance penalty.
Whether you need help selecting the right switch model, port configuration, or licensing, our data center specialists are here to help. Fill out the form and we’ll be in touch within one business day.
A division of BlueAlly
Prominence Tower
3475 Piedmont Road NE
Suite 900, Atlanta, GA 30305
Available Monday–Friday, 7am–6pm PST. We respond to all inquiries within 1 business day.