NVIDIA Spectrum-4 SN5600 64-Port 800GbE AI Ethernet Switch
Flagship 51.2 Tb/s spine for large-scale AI Ethernet fabrics.
We help you choose, configure, and deliver the right system — no obligation.
Configuration at a Glance
Tailored per engagement. Full technical overview below.
Overview
The Spectrum-4 SN5600 is a 2U, 64-port 800GbE open Ethernet switch delivering 51.2 Tb/s of switching capacity for leaf and spine roles in Spectrum-X AI clusters. Each unit is specified, configured, lab-tested, and warranty-backed, sourced through NVIDIA-authorized channels and delivered ready to deploy.
Who This Solution Is For
Business Benefits
Maximum fabric density
Sixty-four 800GbE ports in 2U cut switch count, cabling, and rack space across the spine layer.
Lossless AI throughput
Adaptive routing and RoCEv2 congestion control sustain near-line-rate collective performance for distributed training.
Deployment-ready delivery
Switches arrive configured, tested, and warranty-backed so the fabric stands up on schedule.
Typical Business Use Cases
Spine layer for GPU training clusters
End-of-row aggregation at 800GbE
Spectrum-X lossless Ethernet fabrics
High-bandwidth HPC interconnect
Industry Applications
Technical Overview
Built on the NVIDIA Spectrum-4 ASIC, the SN5600 provides 64 OSFP ports of 800GbE plus a 25GbE SFP28 management port and 33.3 Bpps of packet processing. It runs NVIDIA Cumulus Linux or SONiC and integrates adaptive routing and RoCEv2 telemetry for Spectrum-X AI fabrics.
| Switch ASIC | NVIDIA Spectrum-4 |
| Ports | 64× 800GbE OSFP + 1× 25GbE SFP28 |
| Switching Capacity | 51.2 Tb/s |
| Packet Rate | 33.3 billion packets/sec |
| Fabric Features | RoCEv2, adaptive routing, Spectrum-X |
| Form Factor | 2U, OSFP |
| Network OS | Cumulus Linux / SONiC (open) |
| Power | Redundant hot-swap PSUs, front-to-back airflow |
| Warranty | Manufacturer-backed, authorized channel |
Specifications are indicative and configured to each engagement. Request a quote for a configuration tailored to your requirements.
Warranty, Support & Fulfillment
Every system ships from an authorized channel, configured and tested, with the documentation enterprise buyers need — backed by warranty and a dedicated account team.
Enterprise Warranty
Full manufacturer warranty with optional on-site, next-business-day support and extended coverage.
Authorized Channel
Sourced through Tier-1 distribution and OEM partners — never grey market. Asset & warranty records included.
Lead Time & Deployment
48-hour quotes, then configured, burn-in tested, and delivered on a committed schedule.
Nationwide Fulfillment
Coordinated logistics, rack-and-stack, and delivery wherever your infrastructure lives.
Frequently Asked Questions
How many GPUs can a SN5600 spine support?
A single 51.2 Tb/s SN5600 aggregates dozens of 800GbE leaf uplinks; cluster scale depends on your leaf design, but it comfortably anchors the spine for thousands of GPUs in a Spectrum-X topology. We size the spine count to your target GPU footprint.
Does it require InfiniBand expertise to operate?
No. The SN5600 is standards-based Ethernet running Cumulus Linux or SONiC, so existing Ethernet operations and tooling apply while still delivering RoCEv2 lossless performance for AI.
Can it be split to lower port speeds?
Yes. With splitter cabling the 800GbE ports break out to 400/200/100GbE, letting one platform serve mixed-speed leaf and ToR roles. We configure breakout profiles before delivery.
Hardware Assistance
Configure the NVIDIA Spectrum-4 SN5600 64-Port 800GbE AI Ethernet Switch with Nexus Compute
Tell us your requirements and a hardware specialist will help you specify, configure, and quote the right system — typically within two business days. No obligation.