Cloud Servers with GPU

Use Selectel cloud servers with NVidia® GPUs to perform computation capacity-intensive tasks that require high performance and video RAM bandwidth. GPU-enabled cloud servers will ensure many times faster machine learning, analysis, and graphics processing.

Platform Access for Computing


Used as the dedicated PCI device within the virtual machine. This allows for using all the GPU power and resources.


For ML/DL, inference, video and virtual desktops.

  • Works with most AI frameworks like TensorFlow, PyTorch, Keras, MXNet, Caffe.
  • Compatible with all types of neural networks: GAN, CNN, RNN, SSNN and others.
  • 2560 CUDA cores, 320 NVIDIA Turing Tensor Cores, 16 GB GDDR6.
  • Price per GPU: hourly — ₽40,59; daily — ₽974,16; monthly — ₽29,652,45.

Virtual Machines in the Cloud

  • Intel® Xeon® 2680v4 CPUs,
  • 32 to 128 GB DDR4 RAM

Solve Complex Problems Faster

Machine Learning

GPU is perfect for working with neural networks and deep learning. They facilitate image classification, speech recognition, and big data processing.

Design & Graphics

GPUs boost 3D modeling, video processing, online effect application, and increase the rendering speed manyfold.

High-Performance Computing (HPC)

Engage the GPU in complex computing— e.g. financial analytics, physical process evaluation, and design. Shrink the time to result.

Get on the New Performance Level

Reduce Computing Time

Enhance the performance of your resources: deploy GPU-powered cloud servers within minutes and boost resource-intensive computing.

Pay for Cloud-Based Resources As You Go

Choose the most preferable server configuration and GPU and pay for it only when you need it.

Scale Easily

If the task requires more resources, scale your power up in the Selectel control panel.

Ready to start?