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
Black and silver dual-slot GPU accelerator with green accents, passive heatsink, PCI Express 3.0 x16 edge connector, no fan visible

NVIDIA

NVIDIA M10 32GB GDDR5 graphics card

SKU CH-MN90-TDBPQ MPN M10

A GPU computing processor with four Tesla M10 chips, 32 GB GDDR5 memory, 2650 CUDA cores, PCI Express 3.0 x16 interface, fanless design and ECC memory drawing 225 watt

  • Memory TypeGDDR5
  • 32 GB GDDR5 memory supports large virtualised workloads
  • Four Maxwell GPUs with 2650 CUDA cores accelerate parallel compute tasks
  • Fanless design enables dense server deployments without airflow constraints
  • PCI Express 3.0 x16 interface provides 16-lane bandwidth for GPU computing
  • ECC memory protects data integrity in production environments

Multi-GPU accelerator card

The NVIDIA Tesla M10 packs four Maxwell GPUs on a single PCI Express 3.0 x16 board. Each GPU contributes to a total of 2,650 CUDA cores and 32 GB of GDDR5 memory with ECC support. The fanless design relies on server airflow and draws 225 W under load.

Virtual desktop and compute workloads

This card targets dense VDI deployments and light GPU compute where multiple users share one physical slot. Workloads needing a single large GPU, display outputs, or active cooling should consider other models.

Four GPUs on one board

The defining specification is the four Maxwell graphics processors combined into a dual-slot form factor. That density decides whether the M10 fits your rack density and licensing model.

Highlights

  • 32 GB GDDR5 memory supports large virtualised workloads
  • Four Maxwell GPUs with 2650 CUDA cores accelerate parallel compute tasks
  • Fanless design enables dense server deployments without airflow constraints
  • PCI Express 3.0 x16 interface provides 16-lane bandwidth for GPU computing
  • ECC memory protects data integrity in production environments

Specifications

Memory TypeGDDR5

Questions about this item

Will the M10 fit in my server chassis?

The card uses a PCI Express 3.0 x16 slot and needs an adjacent slot free because it occupies two bracket positions.

What power supply capacity is needed for the 225 watt draw?

Plan for at least 225 watts of dedicated power for the GPU plus the rest of the system; a quality 650 watt or larger PSU is typical for a single-card build.

Can I use the M10 for CUDA development and virtual desktop infrastructure?

Yes, the four GPUs provide 2650 CUDA cores and 32 GB GDDR5 with ECC, making it suitable for parallel compute and VDI workloads.

Also in this aisle

People compared these

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