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
Gold circuit board with silver bracket and PCIe connector

NVIDIA

NVIDIA Tesla V100 16GB HBM2 PCI Express 3.0 graphics card

SKU CH-QM4B-Y8WEA

A GPU with 16 GB HBM2 memory and a PCI Express 3.0 interface for compute workloads

  • 16 GB HBM2 memory feeds large models and datasets
  • PCI Express 3.0 interface fits standard server slots
  • Tesla V100 architecture targets AI and HPC workloads
  • 16 GB capacity reduces multi-GPU scaling needs
  • New condition ensures full factory reliability

Tesla V100 16GB HBM2 PCIe 3.0 Compute Card

This NVIDIA Tesla V100 delivers 16GB of HBM2 memory on a PCI Express 3.0 add-in card. It is built for data-centre workloads such as AI training, high-performance computing and large-scale inference. The board ships in brown-box packaging for system integrators and OEM deployments.

Suits Dense GPU Servers; Avoid for Desktop Graphics

This accelerator fits servers that accept full-height, double-slot PCIe compute cards and provide sufficient auxiliary power. It lacks display outputs and consumer driver support, so it is unsuitable for workstations, gaming PCs or any system that requires video output. Buyers needing graphical display should select a Quadro or GeForce product instead.

16GB HBM2 Memory Capacity

The 16GB HBM2 frame buffer determines the maximum model size and dataset partition that can reside on a single card. Workloads exceeding this capacity require multi-GPU configurations or alternative accelerators with larger memory. Verify your framework and dataset footprint against this limit before purchase.

Highlights

  • 16 GB HBM2 memory feeds large models and datasets
  • PCI Express 3.0 interface fits standard server slots
  • Tesla V100 architecture targets AI and HPC workloads
  • 16 GB capacity reduces multi-GPU scaling needs
  • New condition ensures full factory reliability

Specifications

No published specifications for this item yet.

Questions about this item

Should I choose the 16GB HBM2 version for large model training?

The 16GB HBM2 capacity fits larger datasets and models in memory, reducing the need to swap to system RAM during training runs.

What does the 16GB HBM2 figure mean once the card is installed in a server?

It provides 16GB of high-bandwidth memory directly on the GPU, letting the Tesla V100 hold bigger matrices and batches for faster compute.

What does the PCI Express 3.0 interface cap in practice?

The PCI Express 3.0 link limits host-to-device bandwidth to the lane width of your motherboard slot, typically x16 for full throughput.

Also in this aisle

People compared these

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