Hedgehog Supported NVIDIA Devices

Hedgehog tests, certifies and supports NVIDIA devices in Hedgehog fabrics for your AI training, fine tuning, RAG, inference and general purpose compute use cases. Every supported device runs continuous automated testing in our lab—ensuring the Hedgehog VPC and Gateways your tenants rely on work flawlessly for their AI workloads.

Hedgehog gateway and control software runs on commodity servers you can buy just about anywhere.  We support switches from multiple vendors so you can diversify your supply chain and add elasticity to your hybrid cloud.  

Get Hedgehog
hh-illustration_hero
OCP Member Tier logo _bronze horz cmyk

Hedgehog Gateway Servers

You don't need to spend lots of money on expensive, specialized routers for data center interconnect (DCI).  Hedgehog offers a software-based router called Hedgehog Gateway.  This soft router runs on commodity server hardware equipped with NVIDIA ConnectX-7 SmartNICs.  

hedgehog-gateway-server-1

Hedgehog Control Servers

When you download the Hedgehog software appliance, you are going to run it on a control server.  This controller includes an embedded, fully managed Kubernetes implementation which serves the Hedgehog cloud native API, configures the network operating systems running on your switches, and manages your gateway servers.  In sum your control servers are the brains of your AI cloud network.  

hedgehog control server

NVIDIA Switches

Hedgehog supports a wide variety of Cumulus based NVIDIA Switches of various sizes, bandwidth and port capacities for a variety of roles in an AI Fabric.

nvidia-logo-vert

NVIDIA SN5610

Front-end and back-end AI training on Spectrum-4, at the same 64x800G density as the DS5000.

  • Fabrics:  AI Training, Fine Tuning, RAG, Inference
  • Roles: Leaf, Spine
  • Silicon: NVIDIA Spectrum-4
  • Bandwidth: 51.2 Tb/s
  • Ports: 64x OSFP-800G, 2x SFP28-25G
image-2025-12-7_11-20-5

NVIDIA SN4700

400G spine for inference and general-purpose fabrics; the Spectrum-3 alternative to the DS4000 and Z9332F-ON

  • Fabrics: AI Inference, General Purpose Compute
  • Roles: Spine
  • Silicon: NVIDIA Spectrum-3
  • Bandwidth: 12.8 Tb/s
  • Ports: 32x QSFP-DD-400G
SN4700 front angled

NVIDIA SN4600

High-density 200G leaf or spine: 64 ports in 2U, breaking out to 128x100G.

  • Fabrics:  AI Inference, GPC
  • Roles: Leaf, Spine
  • Silicon: NVIDIA Spectrum-3
  • Bandwidth: 12.8 Tb/s
  • Ports: 64x QSFP56-200G
SN4600 Front

NVIDIA SN3700C

General-purpose 100G leaf or spine; the Spectrum-2 alternative to the DS3000.

  • Fabrics: AI Inference, General Purpose Compute
  • Roles: Leaf, Spine
  • Silicon: NVIDIA Spectrum-2
  • Bandwidth: 3.2 Tb/s
  • Ports: 32x QSFP28-100G
SN3700c

NVIDIA SN2700

Puts existing Spectrum 100G switches to work in a Hedgehog fabric.

  • Fabrics:  GPC
  • Roles: Leaf, Spine
  • Silicon: NVIDIA Spectrum
  • Bandwidth: 3.2 Tb/s
  • Ports: 32x QSFP28-100G
SN2700

NVIDIA SN2200

1G copper top-of-rack for out-of-band management, with 100G uplinks.

  • Fabrics: General Purpose Compute
  • Roles: Management
  • Silicon: NVIDIA Spectrum
  • Bandwidth: 448 Gb/s
  • Ports: 48x RJ45-1G, 4x QSFP28-100G
nvidia-spectrum-sn2000-series-ari