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
HPE V100 16 GB HBM2 dual-slot workstation graphics card

HPE

HPE V100 16 GB HBM2 workstation graphics card

SKU CH-W5GK-TVHHS MPN V100

Workstation GPU with 16 GB HBM2 memory and CUDA cores for accelerated deep learning and HPC workloads

  • 16 GB HBM2 memory for large deep-learning models
  • Thousands of CUDA cores accelerate HPC workloads
  • Dual-slot design fits select HPE ProLiant servers
  • Power-efficient GPU computing for scientific applications
  • New condition ensures full factory reliability

16 GB HBM2 deep-learning accelerator

The HPE V100 is a workstation graphics card built around a 16 GB HBM2 frame buffer and NVIDIA CUDA cores. It plugs into a dual-slot PCIe space and targets high-performance computing, deep learning and demanding 3D simulations. The board delivers GPU compute density for select HPE ProLiant server families.

Suits GPU-heavy workloads, not general graphics

Researchers and engineers running large-scale simulations or training neural networks will use the parallel architecture. IT buyers who only need display output or light CAD work should look at something else. The part also requires a compatible HPE server chassis with adequate power and cooling.

16 GB HBM2 capacity

The 16 GB HBM2 memory size is the single figure that decides fit. Models that exceed this frame buffer will spill to system memory and lose performance. Buyers needing more than 16 GB of device memory must choose a different accelerator.

Highlights

  • 16 GB HBM2 memory for large deep-learning models
  • Thousands of CUDA cores accelerate HPC workloads
  • Dual-slot design fits select HPE ProLiant servers
  • Power-efficient GPU computing for scientific applications
  • New condition ensures full factory reliability

Specifications

No published specifications for this item yet.

Questions about this item

Should I pick the 16 GB version or is there a smaller capacity available?

This card is offered only with 16 GB of HBM2 memory, so there is no lower-capacity alternative to consider.

What does the 100 GB/s bandwidth actually deliver inside a workstation?

The 100 GB/s HBM2 bandwidth feeds the CUDA cores so large deep-learning or HPC datasets move without stalling the compute pipeline.

Which interface does the card use and what throughput ceiling does it impose?

The accelerator installs in a dual-slot PCIe slot, and the PCIe generation of your server caps the host-to-card transfer rate.

Also in this aisle

People compared these

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