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
Full-length full-height PCIe 3.0 card with 16GB HBM2 memory

NVIDIA

NVIDIA Tesla V100 16GB HBM2 PCIe 3.0 workstation graphics card

SKU CH-J0XN-YYVB4 MPN V100

16 GB HBM2 accelerator card for AI training and inference via PCI Express 3.0

  • InterfacePCI Express 3.0
  • Chipset ManufacturerNVIDIA
  • GPUTesla V100
  • Memory Size16GB
  • Memory TypeHBM2
  • 16 GB HBM2 memory accelerates large-model training
  • PCI Express 3.0 interface fits standard workstation slots
  • Full-height full-length form factor suits dense server chassis
  • Tesla V100 GPU targets AI and HPC workloads
  • 3 lb weight eases rackmount installation

AI accelerator for compute workloads

The NVIDIA Tesla V100 is a 16GB workstation graphics card built on the Tesla V100 GPU. It connects via PCI Express 3.0 and uses HBM2 memory to deliver high bandwidth for data-intensive tasks. The full-length, full-height plug-in card weighs 3 lbs and is supplied in a brown box. It targets researchers and engineers running AI training, inference or scientific simulations.

Suited for single-GPU compute.

This card fits users who need one accelerator with 16GB of fast memory for model development or HPC jobs. Buyers who want more memory per GPU, multi-GPU scaling via NVLink, or certified drivers for professional visualisation applications should look at other products. The PCI Express 3.0 interface also limits peer-to-peer bandwidth compared with newer interconnects.

16GB HBM2 memory capacity

The 16GB HBM2 frame buffer is the specification that determines whether this accelerator can hold your working dataset. Models or datasets larger than 16GB will not fit in GPU memory and will require a card with greater capacity or a multi-GPU approach. Capacity is fixed and cannot be expanded.

Highlights

  • 16 GB HBM2 memory accelerates large-model training
  • PCI Express 3.0 interface fits standard workstation slots
  • Full-height full-length form factor suits dense server chassis
  • Tesla V100 GPU targets AI and HPC workloads
  • 3 lb weight eases rackmount installation

Specifications

BrandNVIDIA
ModelTesla V100 16GB
InterfacePCI Express 3.0
Chipset ManufacturerNVIDIA
GPUTesla V100
Memory Size16GB
Memory TypeHBM2

Questions about this item

How much heat does the 16 GB HBM2 card produce during continuous AI workloads?

The Tesla V100 is a full-height, full-length accelerator rated for sustained compute; plan on strong case airflow or rack cooling to handle the thermal output of a 16 GB HBM2 PCIe 3.0 card running all day.

What motherboard slot and power connectors does the PCI Express 3.0 Tesla V100 need?

It requires a full-size PCIe 3.0 x16 slot and the standard auxiliary power cables your workstation PSU provides for a full-height, full-length add-in card.

Will the V100 16 GB HBM2 work in an older chassis that only fits half-length cards?

No — the card is full-length and full-height, so the chassis must accept a full-size plug-in card; older half-length bays will not fit it.

Also in this aisle

People compared these

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