Skip to main content

GPUs

NodeShift GPUs are NodeShift Virtual Machines (VMs) which are flexible and scalable on-demand resources with a wide array of GPUs. GPU-powered Virtual Machines offer greater control over the computing environment compared to other options as you can configure GPUs, CPUs, RAM and Storage as necessary.

With NodeShift GPUs, you can enjoy the benefits of virtualization without the need to purchase or maintain physical hardware. However, it's important to note that virtual machines still require maintenance tasks such as configuration, patching, and software installation.

NodeShift GPUs serve various purposes, such as:

  1. ML/AI Training: empower your work stream with GPUs designed for the latest deep learning techniques, such as neural networks and generative adversarial networks.

  2. Development and testing: They enable the quick creation of a virtual environment with specific configurations for coding and GPU-intensive testing applications.

  3. Cloud-based applications: Running your GPU-consuming applications such as video editing, 3D graphics rendering, and much more on a virtual machine in NodeShift can be cost-effective as you can scale the number of virtual machines based on demand. You pay for additional virtual machines only when needed and can shut them down when not in use.

The scalability of virtual machines allows your application to scale up or out as required to meet your specific needs.

What to consider when creating a GPU VM?

In order to create your GPU deployment, you need to consider the following the options:

  • GPU Type

  • Number of GPUs

  • CPU, RAM and Storage configurations of your VM

  • Total GPU TeraFLOPS

  • Deep Learning Performance Score

  • The operating system

  • Internet Download and Upload speed

What are the available GPU types?

The GPUs that are available for deployment through NodeShift are listed here.

Geographical Availability

GPU VMs are available across the entire world and it is entirely up to you where to create your VMs based on your application’s requirements. In NodeShift, the location of your VM is titled Region and can be selected upon the creation of your GPU VM.

GPU VM Sizes and Pricing

NodeShift provides a diverse range of GPU VM sizes to accommodate various use cases effectively. GPU VM sizes and pricing are tailored to suit your specific workload requirements. The size you select directly impacts crucial factors like processing power, memory, storage capacity, and network bandwidth.

When it comes to pricing, NodeShift adopts a per minute consumption charging model that considers the size of the virtual machine and its configurations. For partial hours, you are billed solely for the minutes used, ensuring fair and precise cost allocation. It's important to note that storage costs are separate and priced accordingly.