docscompute-type
Last edit February 25, 2025

Compute Type

Compute configurations that define the power behind your DiPhyx projects

Compute Type refers to the instance and VM variations offered by providers—including AWS, GCP, and Azure—consisting of two primary factors: CPU cores and memory size.

However, compute types include more configurations: Architecture, either ARM64 or AMD64, Operation System (OS), and the distribution of the OS.

Cloud providers like AWS, GCP, and Azure categorize their compute instances based on several factors, such as CPU, memory, storage, networking capacity, and GPU support. These providers use the following titles for the different types they offer for their instances and VM's:

  • AWS: Instance types. These comprise varying combinations of CPU, memory, storage, and networking capacity and give you the flexibility to choose the appropriate mix of resources for your applications.
  • GCP: Machine types. Each machine type provides a set of resources for your compute instance, such as vCPUS, memory, disks, and GPUs.
  • Azure: VM types. These offer various combinations of CPU, memory, storage, and networking capabilities, allowing you to select the right mix of resources for your applications.

Each provider uses its own naming conventions and offers specific instance types optimized for different workloads.

On DiPhyx, we bring together these variations under the term "Compute Type," which specifies the CPU core count and memory size. You can independently select "Compute Disk" in another field based on your specific requirements.

When creating a compute unit, your choice of provider determines the available compute types you can select. For example, with AWS as your provider, options include compute types like T2 Micro with 1 CPU and 1 GB memory, or C7I 2Xlarge with 8 CPUs and 32 GB memory.

As DiPhyx continues to evolve, additional compute type options will be introduced to better meet user needs.

Defining Compute Types

DiPhyx offers many compute types from each provider in its namespace. However, if you want to add your custom compute types, you need to upgrade to the Enterprise plan and be the owner of a new namespace.