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

NVIDIA

NVIDIA A100 900-21001-0000-000 40GB HBM2 workstation graphics card

SKU CH-0FMC-PS1GR

Delivers 40 GB HBM2 memory on a 5120-bit bus with PCI Express 4.0 x16 for compute workloads

  • InterfacePCI Express 4.0 x16
  • Chipset ManufacturerNVIDIA
  • GPUA100
  • Core ClockBase: 765 MHz Boost: 1410 MHz
  • Memory Clock1215 MHz
  • Memory Size40GB
  • 40 GB HBM2 memory enables massive model training
  • 5120-bit interface delivers extreme bandwidth for compute workloads
  • PCI Express 4.0 x16 provides high-throughput host connectivity
  • Fanless dual-slot design fits dense server configurations
  • 250 W TDP with single 8-pin power simplifies deployment

NVIDIA A100 40GB HBM2 workstation card

The NVIDIA A100 900-21001-0000-000 is a workstation graphics card built for compute-intensive workloads. It carries 40GB of HBM2 memory across a 5120-bit interface and connects through PCI Express 4.0 x16. The fanless, full-height full-length design targets servers and dense GPU nodes where airflow is managed externally.

Check PCIe 4.0 x16 slot

Verify that your motherboard provides a PCI Express 4.0 x16 slot with enough physical clearance for a dual-slot FHFL card. The board draws up to 250W and requires a single 8-pin auxiliary power connector. Ensure your chassis can supply the required power and that the slot is not shared with other high-bandwidth devices.

Fanless dual-slot card needs directed airflow

Because the cooler is fanless, the card relies on strong front-to-back or ducted airflow from the case or server fans. The dual-slot bracket frees adjacent slots for additional cards in multi-GPU configurations. Plan cable routing for the 8-pin connector and confirm that the FHFL length fits your enclosure depth.

Highlights

  • 40 GB HBM2 memory enables massive model training
  • 5120-bit interface delivers extreme bandwidth for compute workloads
  • PCI Express 4.0 x16 provides high-throughput host connectivity
  • Fanless dual-slot design fits dense server configurations
  • 250 W TDP with single 8-pin power simplifies deployment

Specifications

BrandNVIDIA
Model900-21001-0000-000
InterfacePCI Express 4.0 x16
Chipset ManufacturerNVIDIA
GPUA100
Core ClockBase: 765 MHz
Boost: 1410 MHz
Memory Clock1215 MHz
Memory Size40GB
Memory Interface5120-bit
Memory TypeHBM2
CoolerFanless
System RequirementsMax TDP Power: 250W
Auxiliary Power Connector8-pin
Form FactorFHFL
Package ContentsDual-slot

Questions about this item

Does the card ship with any cables, brackets or software licences, or do I need to buy them separately?

The package contains the dual-slot graphics card only; power cables, mounting brackets and any software licences must be sourced separately.

Is the 40 GB HBM2 model the only memory option for the A100 900-21001-0000-000?

This specific part number is the 40 GB HBM2 configuration; other A100 variants with different memory sizes carry different part numbers.

What does the 1410 MHz boost clock mean for actual throughput in a workstation?

The 1410 MHz boost clock is the peak GPU frequency the card can reach under load, limited by the 250 W TDP and thermal conditions in your chassis.

Also in this aisle

People compared these

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