Welcome to OsoDose Store!

Products

NVIDIA A40 48GB

Nvidia

1,00

Size

Delivery will be made from 21 days. Free Shipping for orders $1000+

10 in stock

11
0,20 0,10
Addons Price  :  0,00
Total                :  1,00

NVIDIA A40 48GB. Ampere GPU with 48GB ECC for data centers & Viz. Direct import, 1–3y warranty, compliant docs, shipping 7–10 days across EU.

Product description

Direct import from the US and Europe without intermediaries. Best price in Europe. Official warranty 1-3 years NVIDIA A10 24GB

NVIDIA A10 24GB (Ampere, GA102, GDDR6) is a high-performance professional graphics accelerator that occupies a key position between consumer GeForce cards and flagship A100/H100 series solutions. With 9216 CUDA cores, 288 third-generation Tensor Cores, and 72 RT cores, the model combines the power for artificial intelligence tasks and the flexibility for graphics computing.

The graphics card is based on the modern NVIDIA Ampere architecture using an 8nm process, which provides an optimal balance between energy efficiency and computing performance. Total Tensor performance reaches 312.4.

Detailed specifications of NVIDIA A10 24GB

  • Graphics processor: GA102
  • Process technology: 8 nm
  • CUDA cores: 9216
  • Tensor Cores: 288 (3rd generation)
  • RT cores: 72 (3rd generation)
  • GPU frequency (base): 885 MHz
  • GPU frequency (Boost): 1695 MHz
  • Video memory capacity: 24 GB
  • Memory type: GDDR6
  • Memory frequency: 12500 MHz
  • Bus width: 384 bits
  • Memory bandwidth: 600 GB/s
  • Video encoders: 1× NVENC (9th generation), 2× NVDEC (6th generation)
  • Maximum power (TDP): 150 W

How NVIDIA A10 differs from other models

  • Compared to NVIDIA A40 48GB. A40 offers double the memory (48 GB) and increased FP32/FP64 power, making it suitable for visualisation and heavy graphics tasks. The A10 is a more energy-efficient option (TDP 150 W vs. 300 W), ideal for mass inference and applications where resource savings are important.
  • Compared to NVIDIA A30 24GB. The A30 is equipped with HBM2e memory with higher bandwidth (933 GB/s versus 600 GB/s for the A10), but its peak FP32 performance is lower (10.32 TFLOPS versus 31.24 TFLOPS for the A10). Thus, the A10 will be preferable for graphics and visualisation, while the A30 will be preferable for purely AI computing.
  • Compared to RTX 4500 Ada 24GB. Despite having a similar graphics processor (GA102/AD), the RTX A10 wins out thanks to its ECC memory, vGPU support, and server form factor, which makes it more reliable for round-the-clock operation in data centres.

Where NVIDIA A10 24GB is used

  • AI inference. Processing large data sets and generative models with high speed and stability.
  • 3D graphics and CAD. An excellent choice for engineers, architects, and designers working with modelling and visualisation.
  • Media and video streaming. NVENC/AV1 and NVDEC hardware codecs simplify real-time video processing and transcoding.
  • Virtualisation (vGPU). Ability to share resources to create virtual workstations in corporate systems.

Why you should buy NVIDIA A10 24GB from us

  • Direct import from the US and Europe — no middlemen and no extra markups.
  • 1–3 year warranty from official NVIDIA distributors.
  • Delivery in 7–10 days throughout Europe with full support at every stage.
  • Any form of payment. Card, bank transfer (w/ex VAT), USDT cryptocurrency.
  • Selection for business tasks. We will help you choose and assemble a server solution for machine learning, rendering or virtualisation.

NVIDIA A10 24GB is a reliable and energy-efficient professional-grade graphics card based on the Ampere GA102 architecture. It is ideal for companies looking for a balance between performance and economy, as well as those who want to get an official warranty and the best price in Europe thanks to direct import without intermediaries.

Order the NVIDIA A10 24GB from our online store and accelerate the development of your AI projects and workflows today.

Additional information

Dimensions 267 × 112 cm
Country of manufacture

Taiwan

Manufacturer's warranty (months)

Model

NVIDIA A40

Graphics Processing Unit (Chip)

Technological process

8 nm

Architecture

Ampere

Number of CUDA cores

10752

Number of Tensor cores

336

Number of RT cores

84

GPU Frequency (MHz)

1305

GPU Boost Frequency (MHz)

1740

Video memory size (GB)

48

Memory type

GDDR6

Memory frequency (MHz)

1812

Memory bus width (bits)

384

Memory Bandwidth (GB/s)

695.8

Cache L1 (KB)

128

Cache L2 (MB)

6

Connection interface (PCIe)

PCIe 4.0 x16

FP16 performance (TFLOPS)

37.42

FP32 performance (TFLOPS)

0.5846

FP64 performance (TFLOPS)

0.5846

NVENC / NVDEC number of blocks

1×NVENC, 2×NVDEC

Cooling type

Passive (server module)

Number of occupied slots (pcs)

2

Temperature range (°C)

0–85

Multi-GPU support

Yes, via PCIe/Sync, but without NVLink

Virtualization/MIG support

vGPU support, not MIG

Width (mm)

112

Length (mm)

267

Product reviews

0
0 reviews
0% average rating
5
0
4
0
3
0
2
0
1
0

Reviews

There are no reviews yet.

Only logged in customers who have purchased this product may leave a review.

I found 35 items that matched your query "".