We take cryptocurrency — so the coins you hold can buy real hardware.Pay in crypto, spend it on real hardware. Free insured shipping worldwide over $399. Plain, unbranded boxes, sent from Germany.Free insured shipping over $399. Your rate is locked for 30 minutes once checkout opens.Rate locked 30 minutes. Every item sealed, serial-checked and covered by a 24-month warranty.Sealed, 24-month warranty.

Saved items Sign in
Green rectangular NVLink bridge with four gold connectors

NVIDIA

NVIDIA 900-23946-0000-000 NVLink Bridge 4 Slot server accessory

SKU CH-K7RS-ZYZ9J MPN 900-23946-0000-000

Four-slot NVLink bridge connects H200 NVL GPUs for high-speed GPU-to-GPU communication

  • Enables multi-GPU scaling for H200 NVL servers
  • Four-slot bridge connects up to four accelerators directly
  • High-bandwidth link reduces inter-GPU communication latency
  • Purpose-built for PCIe-based NVIDIA H200 NVL systems

NVLink Bridge 4 Slot for H200 NVL PCIe

This NVIDIA server accessory connects up to four H200 NVL PCIe GPUs in a single system. It is built for administrators who need high-bandwidth GPU-to-GPU communication in dense compute nodes. The part installs directly onto compatible cards without additional cabling. It suits clusters that scale workloads across multiple accelerators.

Four-slot bridge width determines fit

The bridge spans exactly four NVLink fingers on the H200 NVL PCIe board, so the server chassis and motherboard layout must provide matching slot spacing. Systems with fewer than four GPUs or with different slot pitch cannot use this bridge. Verify the platform mechanical drawing before ordering. No other GPU model or slot count is supported.

Highlights

  • Enables multi-GPU scaling for H200 NVL servers
  • Four-slot bridge connects up to four accelerators directly
  • High-bandwidth link reduces inter-GPU communication latency
  • Purpose-built for PCIe-based NVIDIA H200 NVL systems

Specifications

No published specifications for this item yet.

Questions about this item

What does the 4-slot NVLink bridge enable when installed in a server?

It connects up to four H200 NVL PCIe GPUs directly, allowing them to share memory and compute workloads across the GPUs without going through the system CPU or PCIe bus.

Which interconnect standard does the bridge use and what bandwidth limit applies?

The bridge uses NVIDIA NVLink; the practical throughput is capped by the NVLink generation implemented on the H200 NVL PCIe GPUs themselves.

New to this? Buying server hardware with crypto — what to check, what it costs to pay, and what is never asked for.

Also in this aisle

People compared these

Same shelf, same checkout — eight coins and a 30-minute rate lock.