How it works

How it works#

At a very high-level, the functionalities of the API are quite simple. Using simple python scripts from your laptop, you can streamline your entire simulation workflow.

In essence, the API enables you to:

Inductiva API Usage Flow
  1. Start and manage remote Virtual Machines (VMs), operating either independently or collectively as an MPICluster, that are equipped with pre-installed simulation software.

  2. Send your simulation scripts from your laptop to one or more remote machines, and start the simulators.

  3. Download simulation results, selecting either all data files generated or specific ones of interest.

That’s it! While this outlines the basic flow of usage of the API, there are nuances and additional options available, which we’ll explore in detail throughout this section. More specifically, you’ll become familiarized with how the API works:

  • Tasks: Learn about tasks, the API’s core computational object, which gets created when you submit your simulation request, enabling you to track its progress and access its outputs in real-time.

  • Shared and Dedicated Resources: Discover how the API allocates and runs your simulation tasks on computational resources that are either shared among you and other users, or dedicated only to your use.

  • Storage and Data Flow: Get to know the typical flow of data when you invoke a remote simulator using the Inductiva API, from sending your input files to remote computational resources to accessing the simulation outputs.

  • Computational Infrastructure: Take a further look under the hood to learn the underlying infrastructure powering your simulations, especially how the API enables you to access a variety of computational resources, manages your computational workload, and allocates it to the appropriate computational resource through a unified Python code.

  • Templating Engine: Explore how the API enables you to transform fixed parameter values in your “base case” simulation configuration files into programmable variables you can adjust for running large variations of your simulation.

  • Configuring Simulators: Explore how the API wraps existing open-source software packages within layers that facilitate their execution across various cloud-based virtual machines and providers, transforming them into abstract computational tasks.