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 900-2G133-0010-000 L40 48 GB DDR6 AI accelerator card

SKU CH-JVPR-3AFTF MPN 900-2G133-0010-000

Delivers 48 GB of DDR6 ECC memory for AI workloads with a passive dual-slot design

  • 48 GB DDR6 ECC memory enables large-model training
  • Ada architecture accelerates AI inference workloads
  • Passive dual-slot design fits dense server racks
  • ECC protection maintains data integrity during compute

NVIDIA L40 AI accelerator with 48 GB DDR6 ECC

The NVIDIA L40 is a dual-slot passive AI accelerator card built on the Ada architecture. It delivers 48 GB of DDR6 memory with ECC for data integrity in compute workloads. The card is designed for server environments where sustained throughput matters.

Passive dual-slot form factor requires server

This card uses a passive cooler and occupies two slots, so it depends on directed chassis airflow for thermal management. It is not suited for workstations or enclosures without forced-air cooling paths. Buyers needing an actively cooled card or a single-slot profile should consider other options.

How it compares

Measured against everything we stock in GPUs for AI & Data Centers, at today’s prices. Recalculated every night.

Price $5,837.99
Pricier than about 55% of the 36 comparable items (median $5,369.99).

New to this? Buying server hardware with crypto — what to check, what it costs to pay, and what is never asked for.

Specifications

Size & build
Shipping weight0.91 kg
Package size254 × 254 × 254 mm

Questions about this item

How much data can the 48 GB DDR6 ECC memory hold for a single training job?

The 48 GB DDR6 ECC memory stores model weights, activations and datasets up to its full capacity, with error correction protecting every bit during long compute runs.

What does the passive dual-slot design mean for server airflow?

The passive dual-slot card relies on the chassis fans to push air through its heatsink, so the server must provide enough front-to-back flow to keep temperatures in check.

More in GPUs for AI & Data Centers

Similar items